JP2018190070A - Interaction support method, device and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To realize an interaction environment such as a conference capable of reducing synchronization effect of an interaction person by using a computer system.SOLUTION: An interaction support method applies to a system which displays an image showing a person to realize an interaction function, by using a computer system. The method executes: processing which acquires information indicating an interaction state including the number of interaction persons, by an utterance details distribution unit (15); and the processing which allocates utterance of the interaction person to a predetermined number of images on the basis of the information indicating the interaction state.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a dialogue support method, apparatus, and program applied to a system that realizes a dialogue function by displaying an image showing a person.

  In recent years, a system that supports a dialogue between people using a computer system has been proposed. For example, there has been proposed a conference support system having an editing function that displays texts of speech content in a conference and an illustration selected by a user on a display (see Patent Document 1).

  Here, in a dialogue environment such as a meeting, for example, when one option is determined from a plurality of options, a phenomenon of synchronization effect (social synchronization effect) in which the behavior of others influences decision making is known. (See Non-Patent Document 1). For this reason, in a dialog environment such as a meeting, for example, when one option is determined by majority vote from a plurality of options, it is desirable that the synchronization effect on the majority can be reduced.

JP 2017-4270 A

Tsukasa Matsubara and two others, “Study on factors in popularizing electric vehicles considering the effect of synchronization with others in the selection / selection stage”, JSCE Proceedings D3 (Civil Engineering), Vol.68, No.5 (Civil Engineering) Planned Research and Papers Vol. 29), I_691-I_699, 2012

  In the conventional conference support system, for example, in a dialogue environment such as a conference, the synchronization effect for the majority cannot be reduced. Psychologically, the majority can easily listen to the opinions of the minority, and the minority can form a dialogue environment that makes it easier to speak out. Thereby, for example, in a stage before determining one option by majority vote from a plurality of options, there is a usefulness such that a conversation person (conference attendee or the like) can consider each option more evenly.

  SUMMARY OF THE INVENTION An object of the present invention is to realize a dialog environment such as a meeting that can reduce the synchronization effect of a dialog person using a computer system.

  An aspect of the present invention is a dialogue support method applied to a system that realizes a dialogue function by displaying an image showing a person by using a computer system, and acquires information indicating a dialogue state including the number of talkers A dialogue support method for executing a process and a process of assigning a conversationer's speech to a predetermined number of images based on information indicating the conversation state.

  Since it is possible to assign a speech of a conversation person to a predetermined number of images and make the speech speak, it is possible to realize a conversation environment such as a meeting that can reduce the synchronization effect. As a result, for example, even when a choice is decided by a majority decision from a plurality of choices, it is possible to provide a forum for discussion so that each person can consider each choice equally by suppressing the synchronization effect on the majority. Become.

The block diagram which shows the structural example of the dialogue assistance system which concerns on embodiment of this invention. The block diagram which shows the structural example of the computer system which implement | achieves the main function part of the control part which concerns on embodiment of this invention. The flowchart for demonstrating the process of the dialog state grasping | ascertainment part contained in the control part which concerns on embodiment of this invention. The flowchart for demonstrating the process of the number determination part contained in the control part which concerns on embodiment of this invention. The flowchart for demonstrating the process of the number determination part contained in the control part which concerns on embodiment of this invention. The flowchart for demonstrating the process of the number determination part contained in the control part which concerns on embodiment of this invention. The flowchart for demonstrating the process of the statement content distribution part contained in the control part which concerns on embodiment of this invention. The flowchart for demonstrating the process of the statement content distribution part contained in the control part which concerns on embodiment of this invention. The figure for demonstrating the control which switches an avatar in the process which allocates the statement content which concerns on embodiment of this invention to each avatar. The figure for demonstrating the process which collects the comment of many dialog persons in the single avatar in the process which allocates the content of the utterance which concerns on embodiment of this invention to each avatar. The figure for demonstrating the process which collects the speech of many dialog persons in a several avatar in the process which allocates the content of the speech which concerns on embodiment of this invention to each avatar. The block diagram which shows the structural example of the system which concerns on the modification of this invention. The flowchart for demonstrating the process of the system which concerns on the modification of this invention.

Embodiments according to the present invention will be described below with reference to the drawings.
[System configuration]
FIG. 1 is a block diagram showing a configuration of a dialogue support system according to an embodiment of the present invention. As shown in FIG. 1, the system is roughly divided into a dialogue number control system 10 and a user terminal 20.

  Here, this system implements an interactive environment such as a conference using a computer system, and can be applied to a conference system as a representative example. In this embodiment, the dialogue environment includes an environment such as a meeting where each talker can talk to each other, and an environment such as a lecture or an advertisement where a specific talker conveys a comment to a plurality of viewers. .

  The interactive number control system 10 which is a main part of the system is realized by a server, for example, and includes a communication unit 11, a control unit 12, and a sentence end expression database (for convenience, expressed as a sentence end expression DB) 16. Have. The communication unit 11 communicates with the user terminal 20 by a wireless communication method such as Bluetooth (registered trademark), WiFi (registered trademark), or infrared communication, and transmits and receives various types of information.

  The user terminal 20 is a personal computer (PC) or a smartphone. In the present embodiment, the user terminal 20 means a plurality of user terminals 20-1 to 20-n operated by each user corresponding to each dialoguer.

  The control unit 12 includes a dialogue state grasping unit 13, a number-of-person determining unit 14, and a statement content distributing unit 15 as main function units of the system 10. As shown in FIG. 2, the control unit 12 includes a computer system having a processor (CPU) 120, a memory 121, an input unit 122, and an output unit 123. The main function units 13, 14, and 15 of the control unit 12 are realized by the CPU 120 executing software set in the memory 121.

  The dialog state grasping unit 13 transmits, via the communication unit 11, information such as a plurality of options transmitted from the user terminal 20 that indicate items to be determined in a meeting, for example, and options that each dialoger (user) agrees with. Receive. Based on the received information, the dialogue state grasping unit 13 grasps (calculates) the total number of conversations participating in the dialogue (meeting), the number of adults of each choice, and the choices agreed with by each dialogue, and stores the memory. 121 is a functional unit stored in 121. The dialog state grasping unit 13 may have a function of inputting all information to be grasped from the user terminal 20 and storing it in the memory 121 without executing the grasping process.

  The number-of-persons determination unit 14 obtains information indicating the total number of conversations, the number of adults for each choice, and the choices approved by each talker from the dialogue state grasping part 13, and responds to each choice based on this information. This is a functional unit that determines the number of avatars (described later).

  The message content distribution unit 15 acquires the number of avatars corresponding to each option from the number-of-people determination unit 14, and transmits from the user terminal 20 based on the number of avatars and the above-described information received from the dialogue state grasping unit 13. It is a functional unit that determines which avatar is to speak the content of the comment to be made. The user terminal 20 transmits the content of each conversation person's speech to the communication unit 11 as voice data or text data.

  In other words, the statement content distribution unit 15 distributes the statement content to each avatar corresponding to the conversation person (speaker) based on the conversation contents of the conversation person. As a result, for example, a single dialoguer who is a minority speaks with multiple dialoguers (avatars), and conversely, a small number of dialoguers (avatars) speak with the majority. In addition, an apparent dialogue environment can be realized.

  As will be described later, the sentence end expression DB 16 is a storage that accumulates information to be referred to when switching processing of an avatar to be spoken by the comment content distribution unit 15.

Here, the dialogue support system of the present embodiment draws an avatar that is an image showing a person (character) on the display of the user terminal 20, for example, and realizes an interactive environment via the avatar. The avatar may be a 3D image, for example, or may include an animation linked to the voice of the conversation person. The avatar drawing is realized by an application (existing software tool) of the user terminal 20 operated by each user. In addition, when drawing an avatar that includes animation linked to the voice of the interlocutor, it is realized by existing software tools, and further, for example, a voice changer filter or the like cannot be used to identify the interlocutor personally. It is desirable.
[Operational effects of this embodiment]
Next, with reference to FIG. 3 to FIG. 11, the operation of the control unit 12 will be mainly described in the dialogue number control system 10 of the present embodiment.

  In the present embodiment, as a specific example, this system 10 is applied in an interactive environment that corresponds to a conference via an avatar and determines one option from a plurality of options as a matter to be determined by majority vote. In this case, the system 10 may be used from the stage of executing the majority vote in the dialogue through the avatar, or after the majority and the minority are clarified.

  FIG. 3 is a flowchart for explaining the processing of the dialog state grasping unit 13 included in the control unit 12. Here, the communication part 11 of this system 10 will transmit / receive various information regarding the dialogue (conference) via an avatar, if wireless communication is established between the user terminals 20.

  As illustrated in FIG. 3, the dialog state grasping unit 13 of the system 10 inputs information transmitted from the user terminal 20 via the communication unit 11 (YES in Step S1). Here, when receiving a reset command from the communication unit 11 as an initial process for starting the dialogue, the dialogue state grasping unit 13 initializes a memory (memory 121 in FIG. 2) for storing (registering) information. Execute (YES in step S2, S3).

  If it is not a reset command (NO in step S2), the dialogue state grasping unit 13 requests the communication unit 11 to input information until all necessary information for grasping the dialogue state is registered in the memory. (NO in step S4, S7). Specifically, the conversation state grasping unit 13 keeps the communication unit 11 until the information transmitted from the user terminal 20 acquires information indicating a plurality of options and options approved by each dialoger (user). It requests the input of information (step S7).

  When all the necessary information is acquired (registered in the memory), the conversation state grasping unit 13 calculates the total number of conversations participating in the conversation (conference) and the number of adults for each option based on the information ( YES in step S4, S5). Furthermore, the dialog state grasping unit 13 determines the number of information indicating the calculated total number of dialogs, the number of adults for each option, and the options transmitted from the user terminal 20 that each dialoger (user) agrees with. It transmits to the part 14 and the statement content distribution part 15 (step S6).

  4 to 6 are flowcharts for explaining processing of the number-of-people determining unit 14 included in the control unit 12. The number determination unit 14 determines the number of avatars corresponding to each option based on the information acquired from the dialogue state grasping unit 13.

  FIG. 4 is a flowchart for explaining a process in the case where the same number of avatars are associated with each option. As shown in FIG. 4, the number-of-persons determining unit 14 has acquired from the dialog state grasping unit 13 information indicating the total number of dialogs, the number of adults for each option, and the options that each dialoger agrees with. Is determined (step S10). When acquiring the information, the number-of-people determining unit 14 executes processing for determining the number of avatars (YES in step S10).

  The number determination unit 14 sets the same number of avatars for each option as the number of avatars corresponding to each option based on the acquired information (step S11). In this case, the number of avatars may be matched to the majority or minority. The number determination unit 14 calculates the majority or minority number of each option from the information on the total number of conversations and the number of adults of each option. The number-of-people determining unit 14 may set the number of avatars corresponding to each option (all the same number) at random. In addition, the number-of-people determining unit 14 may set the number of avatars determined by the user in accordance with prior input from the user terminal 20.

  After the determination, the number-of-people determining unit 14 transmits information indicating the number of avatars corresponding to each option (here, the same number) to the message content distributing unit 15 and the communication unit 11 (step S12).

  FIG. 5 is a flowchart for explaining a process in the case where an approver associates two avatars with one option in each option. As shown in FIG. 5, the number-of-persons determining unit 14 has acquired from the dialog state grasping unit 13 information indicating the total number of dialogs, the number of adults for each option, and the options that each dialoger agrees with. Is determined (step S20). Upon obtaining the information, the number-of-people determining unit 14 determines whether or not there is an option for one of the supporters based on the information (YES in step S20, S21).

  When there is no such option as the determination result, the number-of-people determining unit 14 executes a process of setting the number of avatars of all options to the same number as the number of adults of each option (NO in step S21, S24). On the other hand, if there is the option as a determination result, the number-of-people determining unit 14 sets the number of avatars to two for only one option of the supporter, and sets the number of avatars for each remaining option to the number of adults The process which sets to the same number is performed (YES of step S21, S22).

  After the determination, the number-of-persons determination unit 14 transmits information indicating the number of avatars corresponding to each option (here, two for only one option to the supporter) to the message content distribution unit 15 and the communication unit 11 (step S1). S23).

  FIG. 6 is a flowchart for explaining processing when the user sets the number of avatars for each option. As shown in FIG. 6, the number-of-people determining unit 14 has acquired the set value of the number of avatars corresponding to each option set by the user before acquiring the above-described information from the conversation state grasping unit 13. Is determined (step S30).

  Next, when the information is acquired from the dialog state grasping unit 13, the number-of-people determining unit 14 executes a process of setting the number of avatars of each option to a number according to the user designation based on the acquired setting value (step S31). YES, S32). After the determination, the number-of-people determining unit 14 transmits information indicating the number of avatars (here, the number designated by the user) corresponding to each option to the message content distributing unit 15 and the communication unit 11 (step S33).

  FIGS. 7 and 8 are flowcharts for explaining the processing of the statement content distribution unit 15 included in the control unit 12. The comment content distribution unit 15 causes the avatar to speak the content of each conversation person based on the information acquired from the dialog state grasping unit 13 and the information indicating the number of avatars corresponding to each option acquired from the number of people determination unit 14. To decide.

  FIG. 7 is a flowchart for explaining the overall processing of the statement content distribution unit 15. FIG. 8 is a flowchart for explaining the allocation process to the avatar shown in step S45 in the process of the message content distribution unit 15.

  As shown in FIG. 7, the statement content distribution unit 15 determines whether or not the information indicating the number of avatars corresponding to each option is acquired from the dialog state grasping unit 13 and the number of avatars corresponding to each option (Step S40). . The information from the dialogue state grasping unit 13 is information indicating the total number of dialogues, the number of adults for each option, and the options that each dialoguer agrees with.

  When the comment content distribution unit 15 acquires the information, it determines an avatar corresponding to each conversation person based on the information (YES in step S40, S41). Here, as a method for determining an avatar, when the number of interlocutors who agree with each option (the number of adults) matches the number of avatars corresponding thereto, a pair of interlocutors and avatars is determined. However, the present invention is not limited to this, and the avatar may be determined at random or may be determined by the user's designation corresponding to each of the interlocutors.

  If the number of interlocutors who agree with each option (the number of adults) is less than the number of corresponding avatars, the remark content distribution unit 15 does not set a specific pair and remarks as described later. Processing to divide the contents into each avatar is executed. On the other hand, if the number of interlocutors who are in favor of each option is greater than the number of corresponding avatars, the contents of the statements are aggregated to the specified avatar without setting a specific pair, as will be described later. Processing will be executed. Furthermore, when the number of interlocutors who agree with some options (the number of adults) matches the number of avatars corresponding thereto, the comment content distribution unit 15 determines some pairs, As will be described later, processing is performed to divide the content of the rest of the talker into each avatar.

  Here, in a dialog (conference) environment via an avatar assumed in the present embodiment, the user terminal 20 transmits the content of each conversation person's speech to the communication unit 11 as voice data or text data. In this embodiment, for the sake of convenience, it is assumed that the content of each conversation person's speech is transmitted to the communication unit 11 by voice data. In the case of audio data, the message content distribution unit 15 executes processing in units of time such as a message time, as will be described later. On the other hand, in the case of text data, processing in which the time unit is replaced with the number of characters is executed.

  Returning to FIG. 7, the message content distribution unit 15 determines whether or not the message content of each conversation person has been received from the communication unit 11 (step S <b> 42). As will be described later, the comment content distribution unit 15 measures the interruption time (Tpn-1) of each conversation person and executes avatar switching processing based on this time (YES in step S42, S43). Here, the message content distribution unit 15 measures the interruption time (Tpn-1) from the last time received from the same talker and the time received this time.

  Here, the speech content distribution unit 15 divides the received speech content into each avatar based on the determination of the avatar corresponding to each conversation person by the processing of step S41 described above (for convenience, the distribution processing 1 and the distribution processing 1). A process to be aggregated (conveniently distributed process 2), or a process that an avatar corresponding to the talker speaks as it is (conveniently distributed process 3) is determined. (Steps S44, S47).

  In the case of the distribution process 1 or the distribution process 2, the message content distribution unit 15 executes a process of assigning the message content to each avatar as described later (YES in step S44, YES in S47, S45). On the other hand, in the case of the distribution process 3, the comment content distribution unit 15 executes a process in which the avatar corresponding to the talker who is the transmission source speaks as it is (NO in step S47, S48).

  The message content distribution unit 15 sends information for designating an avatar determined by the process according to any of the distribution processes 1, 2, and 3, that is, an avatar that speaks the message content of each conversation person, via the communication unit 11. The user terminal 20 is notified (step S46). As a result, in an apparent conversation environment (conference) via an avatar drawn on the display, the avatar that specifies the content of each conversation person speaks.

  Next, with reference to the flowchart of FIG. 8 and FIG. 9, the process of assigning the content of a statement to each avatar in the case of the distribution process 1 will be described. The distribution process 1 is a process in which one speaker's speech is divided into a plurality of avatars and each avatar speaks.

  As shown in FIG. 8, the speech content distribution unit 15 determines whether the speech time (Tsn) of the received speech content is 0 hour or more and less than a predetermined reference value (for example, Tsth = 3000 [msec] here). Is determined (step S50). As will be described later, this determination result is used for the control for switching the avatar when the speech of a single talker is divided into a plurality of avatars. FIG. 9 is a diagram specifically explaining the control for switching the avatar.

  As shown in FIG. 9, it is possible to divide the content of one conversation person's speech into a plurality of speech (1-n) based on speech time (Tsn) and speech break time (Tpn-1). For example, the avatar can be switched so that the utterance (1) is distributed to the avatar (1) and the utterance (2) is distributed to the avatar (2). Here, in the control for switching the avatar, as will be described later, the speech break time (Tpn-1) is compared with the break time threshold value (Tpth ′) calculated based on a predetermined rule. The avatar to be switched is controlled or the current avatar is continuously spoken.

  Returning to FIG. 8, if the determination result in step S50 is YES, the message content distribution unit 15 sets the parameter b to “b = 1” (step S51). This parameter b is used to calculate the above-mentioned break time threshold (Tpth ′). If the determination result in step S50 is NO, the message content distribution unit 15 further determines whether the message time (Tsn) is within a predetermined reference range (for example, 3000 or more and 5000 [msec] or less). Determination is made (step S59). If the determination result in step S59 is YES, the message content distribution unit 15 sets the parameter b to “b = −Tsn / 4000 + 1.75”, for example (step S60). If the determination result in step S59 is NO, the message content distribution unit 15 sets the parameter b to “b = 0.5”, for example (step S60).

  Here, the value of the parameter b is an example, and is set so that the interruption time threshold (Tpth ′) monotonously decreases as the speech time (Tsn) increases within the range of “b ′ ≦ b ≦ 1”. . That is, as the speech time (Tsn) increases, the threshold for switching speech (Tpth ′) decreases. The value of b ′ is set by the user in the range of “0 <b ′ <1”.

  Next, the statement content distribution unit 15 determines whether a so-called sentence ending expression is included in the statement content of one conversation person (step S52). Specifically, the statement content distribution unit 15 refers to the sentence end expression DB 16 and determines whether or not a registered character string is included at the end of the statement content. The registered character string is, for example, one that is used as the ending of “I think” or “I think”. When the content of speech is speech data, the determination is executed after the speech data is converted into text data using a speech recognition system realized by an existing software tool.

  If the utterance is included at the end of the utterance content, the utterance content distribution unit 15 determines that there is a sentence ending expression, and sets the parameter a to “a = 1”, for example (YES in step S52, S53). On the other hand, if it is determined that there is no sentence end expression, for example, the parameter a is set to “a = 2” (NO in step S52, S54). This parameter a together with the parameter b is used to calculate a break time threshold (Tpth ′) as described below.

  The statement content distribution unit 15 calculates the interruption time threshold (Tpth ′) described above from a predetermined calculation formula “Tpth ′ = a × b × Tpth” (step S55). Here, “Tpth” is a natural break time in human speech, for example, “Tpth = 389 [msec]” here. The “Tpth” may be set by the user.

  Next, the message content distribution unit 15 executes avatar switching control based on the result of comparing the message interruption time (Tpn-1) and the calculated interruption time threshold (Tpth ′) (step S56). . Specifically, the speech content distribution unit 15 executes processing for switching the avatar to be speeched if the speech break time (Tpn-1) exceeds the break time threshold (Tpth ′) (YES in step S56). S57). For example, as shown in FIG. 9, if the interruption time (Tp1) from the conversationer's statement 1 to statement 2 exceeds the interruption time threshold (Tpth ′), the message 2 is changed from avatar 1 to avatar 2 I will let you.

  On the other hand, if the speech interruption time (Tpn-1) is less than the interruption time threshold (Tpth ′), the speech content distribution unit 15 executes a process of continuously speaking the current avatar (NO in step S56, S58). For example, as shown in FIG. 9, if the break time (Tp1) from the talker's speech 1 to speech 2 is short due to the break time threshold (Tpth '), the avatar 1 continues to speak 2 It will be.

  Next, with reference to FIG. 10 and FIG. 11, in the case of the distribution process 2, a process for assigning the content of a statement to each avatar will be described. The distribution process 2 is a process in which a relatively large number of dialoguers are aggregated into a small number of avatars.

  As described above, the statement content distribution unit 15 is based on the information acquired from the dialogue state grasping unit 13 and the number-of-people determining unit 14, and the number of dialogues in favor of each option is greater than the number of corresponding avatars. In this case, a process of collecting the contents of the statements into the designated avatar is executed (see step S47 in FIG. 7). Here, information indicating the number of avatars corresponding to each option is acquired from the number-of-people determining unit 14. Further, the dialogue state grasping unit 13 obtains information indicating the total number of dialogues, the number of adults for each choice, and the choices approved by each dialoguer.

  FIG. 10 shows a process of aggregating, for example, the comments of two interactors into a single avatar. That is, the message content distribution unit 15 assigns the message (1) of the conversation person (1) to a single avatar (which may be designated by the user) and causes the avatar to speak. Further, the statement content distribution unit 15 assigns the statement (2) of the conversation person (2) to the same avatar and causes the avatar to speak. Here, in the utterances (1) and (2), there is a time interval corresponding to the utterance interruption time described above.

  FIG. 11 shows a process of aggregating the comments of, for example, three interrogators (1), (2), and (3) into two avatars (1) and (2). First, the statement content distribution unit 15 selects the remaining number (1) of the conversation persons (3) obtained by subtracting the number of avatars (2) from the number of conversation persons (three). The selected conversation person (3) may be randomly selected or specified by the user.

  Next, the message content distribution unit 15 distributes the messages (1) and (2) of the selected dialogue person (3) to the avatars (1) and (2) as shown in FIG. Execute the process. In addition, since the number of avatars (2) coincides with the two conversation persons, the remark content distribution unit 15 and the pair of the conversation persons (1) and the avatar (1) and the conversation persons (2) and the avatar (2 ) And a pair. The pair setting may be random or specified by the user. The statement content distribution unit 15 executes a process of causing the conversations of each of the interlocutors (1) and (2) to speak by a pair of avatars (1) and (2).

  As described above, according to the present embodiment, in a system that realizes an apparent dialogue environment via an avatar (an image showing a person), when a dialoguer's speech is assigned to the avatar and the avatar speaks, the dialogue is performed. The number of avatars corresponding to the utterance can be controlled based on the information indicating the state. In the present embodiment, the information indicating the conversation state specifically indicates the total number of conversations, the number of adults of each option, and the options that each of the interactors can agree with, which can be acquired from the dialog state grasping unit 13. Information.

  Specifically, this embodiment is effective when a dialog environment (conference) in which one option is determined by majority vote as a matter to be determined is applied to a system that implements via an avatar. is there. That is, according to this embodiment, when performing a majority decision, the synchronization effect by the majority can be reduced by controlling the number of avatars corresponding to the statements.

  For example, it is possible to distribute the remarks of a single dialogue person who is a minority to a plurality of avatars. Conversely, the voices of the majority can be distributed to a small number of avatars. As a result, by changing the apparent number of dialogues to be equal, it is possible to realize a dialogue environment that can reduce the synchronization effect on the majority and fully consider the opinions of minorities. In other words, by showing the same number of people in the majority and minority groups, an apparent dialogue environment is created that reduces the synchronization effect of the majority and creates a situation in which minority opinions can be easily examined psychologically. it can. This makes it possible to more accurately consider the advantages and disadvantages of each option during discussions such as meetings.

Here, in the present embodiment, when the conversation content is distributed to a plurality of avatars by the function of the speech content distribution unit 15, based on information such as speech time, speech break time, presence / absence of sentence ending expression, Control of avatar switching and continuation to speak is realized. Therefore, in the present embodiment, it is possible to automatically perform a process of distributing a conversation person's speech to a plurality of avatars.
[Modification]
FIG. 12 is a block diagram showing a configuration of a system according to a modification of the present embodiment. FIG. 13 is a flowchart for explaining processing of the system according to the modification.

  As illustrated in FIG. 12, the present modification relates to a message division system 30 that includes only the message content distribution unit 15 and the sentence ending expression DB 16 of the present embodiment. In addition, this modification is the structure which has the digital signage (digital signage) 40 (it means several digital signage 40-1-40-n) besides the user terminal 20 of this embodiment.

  The message division system 30 of the present modification is a system that makes it appear as if a plurality of avatars are speaking, for example, a single user (corresponding to a conversation person). Specifically, as shown in FIG. 13, the statement content distribution unit 15 acquires the number of avatars transmitted from the user terminal 20 via the communication unit 11 (step S70).

  The statement content distribution unit 15 executes processing for assigning the statement content to each avatar by the same processing (see FIG. 8) as in the above-described embodiment. FIG. 13 shows processing (steps S71 to S82) corresponding to steps S50 to S61 in FIG. 8, and detailed description thereof is omitted. However, in the process of step S78, the switching of the avatar to be spoken may be set at random or in the order set in advance by the user.

  As an application field of this modification, for example, it is possible to display a plurality of avatars on the display of the digital signage 40 installed in an actual store and distribute the remarks of a single store clerk to each avatar. . As a result, it is possible to increase the effects of sales and the like by making it appear as if multiple salesclerks are selling.

  As another application field of this variation, in a class (or lecture) using a large display, multiple avatars are displayed on the display, and one speaker's remarks are distributed to multiple avatars. It is possible to make it. This has the effect of making it easier for the students and viewers to attract attention by making it appear as if a plurality of instructors are taking classes.

  In the present embodiment and the modification, the control of the number of pseudo talkers in the interactive environment using images and animation has been described. However, the present invention is not limited to this, and only audio data is used, for example, for a telephone conference. Is also applicable. In this case, it is possible to control the number of interlocutors using a general-purpose voice changer (such as Morph VoX Pro).

  In short, the present invention is not limited to the above-described embodiment as it is, and can be embodied by modifying the constituent elements without departing from the scope of the invention in the implementation stage. Further, various inventions can be formed by appropriately combining a plurality of constituent elements disclosed in the embodiment. For example, some components may be deleted from all the components shown in the embodiment. Furthermore, you may combine suitably the component covering different embodiment.

10 ... Dialogue number control system, 11 ... Communication part, 12 ... Control part,
13 ... Dialog state grasping part, 14 ... Number of persons determination part, 15 ... Statement content distribution part,
16 ... sentence ending expression database (sentence ending expression DB), 20 ... user terminal,
DESCRIPTION OF SYMBOLS 120 ... Processor (CPU), 121 ... Memory, 122 ... Input part, 123 ... Output part.

Claims (8)

A dialogue support method applied to a system that uses a computer system to display an image showing a person and realize a dialogue function,
A process of obtaining information indicating a conversation state including the number of conversation persons;
A process of assigning a dialoguer's speech to a predetermined number of the images based on the information indicating the dialogue state;
Dialogue support method to execute.   The dialogue support method according to claim 1, further comprising a process of determining the number of images to be used based on information indicating the dialogue state. In the assigning process,
The dialogue support method according to claim 1 or 2, wherein when utterances of a conversation person are assigned to a plurality of images, a process of assignment is performed based on at least information indicating a break time of the utterances. In the assigning process,
The dialogue support according to claim 1 or 2, wherein when utterances of a conversation person are assigned to a plurality of images, a process of assignment is executed based on at least the speech time, the break time of the speech, and information indicating presence / absence of a sentence end expression. Method. In the assigning process,
The process of assigning a plurality of utterances to the same image or a process of aggregating the utterances into a plurality of images when assigning each utterance of a relatively large number of dialogues to a relatively small number of images is performed. The dialogue support method according to any one of the above. In the process of aggregation,
The process of setting up the same number of images as each conversation person and assigning each conversation person's speech to the pair images,
A process of assigning a conversationer's remarks excluded from the pair setting target to a plurality of images;
6. The dialogue support method according to claim 5, further comprising: A dialogue support device applied to a system that uses a computer system to display an image showing a person and realize a dialogue function,
Means for obtaining information indicating a conversation state including the number of conversation persons;
Means for assigning a dialoguer's speech to a predetermined number of images based on the information indicating the dialogue state;
A dialogue support apparatus comprising: A program applied to a system that uses a computer system to display an image showing a person and realize an interactive function,
A procedure for obtaining information indicating the conversation state including the number of the conversation persons,
Assigning a dialoguer's speech to a predetermined number of the images based on the information indicating the dialogue state;
A program that causes a computer to execute.