WO2024023901A1

WO2024023901A1 - Communication terminal, comment output method, and program

Info

Publication number: WO2024023901A1
Application number: PCT/JP2022/028670
Authority: WO
Inventors: 陽子石井; 桃子中谷; 晴美齋藤
Original assignee: 日本電信電話株式会社
Priority date: 2022-07-25
Filing date: 2022-07-25
Publication date: 2024-02-01

Abstract

The purpose of the present disclosure is to output a comment proposed by a person at a suitable timing in the flow of a dialogue.　To this end, the present disclosure is a communication terminal that outputs a comment during a dialogue between a plurality of participants, the communication terminal including: a topic input unit that converts, into text data, speech data that indicates the content of a dialogue that includes a topic of the dialogue; a topic determining unit that determines the content of the topic on the basis of a term that has appeared at least a prescribed number of times in a prescribed time period in the text data; a comment selecting unit that acquires data about a comment pertaining to the determined content of the topic from a storage unit in which the data about the comment proposed by a person has been previously stored, thereby selecting a prescribed comment to be outputted; and an output unit that outputs the data about the selected prescribed comment.

Description

Communication terminal, comment output method, and program

The present disclosure relates to a technology that automatically outputs comments such as answers to topics in a dialog between multiple people.

In a dialogue situation where participants are discussing anecdotes, coming up with ideas, and discussing communication, outputting the opinions of people other than the participants in the discussion can liven up the discussion and increase the excitement. Since it is possible to suppress biased decision-making by participants, it is possible to have meaningful discussions.

Therefore, a technique has been proposed in the past in which a computer automatically generates or selects an appropriate answer sentence for a topic sentence (see Patent Document 1). Furthermore, a technique has been proposed in which a computer generates natural utterance candidates for user's utterances (see Patent Document 2).

JP 2020-4224 Publication JP2015-79383A

However, unlike when a person explicitly asks a computer a question and requests an answer from the computer, in a dialogue between multiple participants, the computer responds at an appropriate time during the flow of the dialogue. It is difficult to output comments. Additionally, if participants in a discussion recognize that the computer is the source of their comments, priority will be given to dialogue between participants, and participants will not listen to the opinions expressed by the computer. Alternatively, there is a problem that there is a tendency not to adopt it as an opinion.

In order to solve the above-mentioned problems, the present invention aims to output comments devised by a person at an appropriate timing in the flow of dialogue.

In order to solve the above problem, the invention according to claim 1 provides a communication terminal that outputs comments during a dialogue between a plurality of participants, and which outputs a text from audio data indicating the content of the dialogue including a topic during the dialogue. a topic input unit that converts the data into data; a topic determination unit that determines the content of the topic based on words that appear more than a predetermined number of times in a predetermined time in the text data; and a topic determination unit that stores comment data devised by a person in advance. a comment selection unit that selects a predetermined comment to be output by acquiring comment data related to the content of the determined topic from a storage unit that has been previously stored; The communication terminal has an output unit that outputs data.

As explained above, according to the present invention, it is possible to output comments devised by a person at an appropriate timing in the flow of dialogue.

1 is an overall configuration diagram of a communication system according to this embodiment. It is an image diagram of dialogue at base α. FIG. 2 is an electrical hardware configuration diagram of each communication terminal and server according to the present embodiment. 3 is a functional configuration diagram of a communication terminal 5. FIG. 5 is a diagram showing processing or operation of the communication terminal 5. FIG.

Hereinafter, embodiments of the present invention will be described based on the drawings. Note that the communication terminal 5 according to the present embodiment provides specific improvements over the conventional technology, and the present embodiment automatically outputs comments such as answers to topics in response to conversations between multiple people. It shows the improvement in the technical field.

[Overall system configuration]
FIG. 1 is an overall configuration diagram of a communication system according to an embodiment.

As shown in FIG. 1, the communication system 1 of this embodiment is constructed by a display device 2, a communication terminal 5, an input/output device 6, a server 7, a communication terminal 8, and a communication terminal 9. The display device 2, the communication terminal 5, and the input/output device 6 are used at the base α (first base) during the dialogue. The communication terminal 8 is being used at the base β (second base) during the dialogue at the base α. The communication terminal 9 is used by an arbitrary person among an unspecified number of people before a conversation at the base α. The server 7 is installed on a cloud or the like, and stores comments sent from the communication terminal 9 before a conversation at the base α, or comments sent from the communication terminal at the base β during a conversation at the base α. I remember the comments I receive. Further, the server 7 transmits the stored comments to the communication terminal 5 during the dialogue at the base α.

The communication terminals 5, 8, and 9 are PCs (Personal Computers) or the like, and are capable of communicating via a communication network 100 such as the Internet.

The display device 2 is a display or the like.

Furthermore, although two bases are shown in FIG. 1, it is also possible to communicate with three or more bases. In this case, the number of communication terminals increases according to the number of bases.

The input/output device 6 is a device that transmits video data obtained by photographing the surroundings and sound data obtained by collecting surrounding sounds to the communication terminal 5. This transmission method may be wired or wireless.

Furthermore, at the base α, the communication terminal 5 acquires the video data and sound data output from the input/output device 6, and transmits the video data and sound data to the communication terminal 10 at the base β. Similarly, the communication terminal 8 at the base β may acquire the video data and sound data from the base β and transmit it to the communication terminal 5 at the base α. Note that the communication terminal 5 will be explained in detail later using FIG. 4.

<Server 7>
The server 7 plays a role as a DB (Data Base) server. Before the dialogue at the base α, information such as comments is obtained from the communication terminal 9 and stored in association with the comments, identification information, and, as the case may be, priorities. Identification information and priority will be explained later.

Also, during the conversation at the base α, information such as comments is acquired from the communication terminal 8 and stored in association with the comments, identification information, and in some cases priorities. Additionally, information such as comments is sent to the communication terminal 5 in response to a request from the communication terminal 5 during the dialogue.

<Communication terminal 8>
The communication terminal 8 at the base β is a PC or the like used by an observer who does not make a direct voice statement during the dialogue at the base α, but who views the video and audio of the dialogue at the base α. By using the communication terminal 8, the observer at the base β can grasp the flow of the dialogue at the base α and the state of the participants in the current dialogue. An output device and an input device are connected to the communication terminal 8.

Examples of the output device include a headphone-type speaker that reproduces the audio of the dialogue, a display that plays a video image that allows you to see the dialogue, a display that reproduces text information of the content of the dialogue, and a display that reproduces the text of the dialogue. These are headphones or speakers that read out the converted information.

Further, the input device is, for example, a microphone that allows voice input, a keyboard that allows character input, or the like. When the observer performs voice input, the communication terminal 8 converts the voice of the input comment into text information.

Furthermore, the communication terminal 8 gives identification information for classifying the content of the comment to the comment obtained by the observer. For example, if the topic is asking about a place, the viewer is asked to input latitude and longitude information on a map, and the latitude and longitude are stored in the server 7 as identification information. Alternatively, the viewer may be asked to input tags such as parks, commercial facilities, cultural facilities, etc. from the content of the comments, and these may be used as identification information. In addition to this, the observer may be asked to input arbitrary information and these may be used as identification information. These identification information may be input by a third party who is not an observer.

Additionally, the identification information may be coordinate values. In this case, the server 7 converts the collected comments into high-order coordinates using APIs (Application Programming Interfaces) such as doc2vec and fast2text, and performs principal component analysis on these coordinates to reduce the dimension to two or three dimensions. Then, the converted coordinates are used as identification information. Note that the server 7 may calculate all the Euclidean distances for each coordinate of the aggregated comments, cluster those with close distances, categorize them based on the clustered comments, and use each category as identification information. The category in this case may be the coordinates of the center of the group after clustering.

Furthermore, a priority order (such as a numerical value) in which comments are used in dialogue may be added to the identification information. Priority is assigned by an observer or third party. The higher the priority, the higher the possibility of appearing in the dialogue, so if the observer has a strong desire to convey the message, the priority is set to "1," for example. There is no particular upper limit to the numerical value of the priority order, and any numerical value can be set. As a result, the server 7 stores the comment, identification information, and/or priority in association with each other.

<Communication terminal 9>
The communication terminal 9 is a PC or the like used by any person (respondent) among the unspecified number of people who are interested in the content of the dialogue or those who meet predetermined conditions before the dialogue is held at the base α. . This respondent uses the communication terminal 9 to transmit and save his/her comments to the server 7 before the conversation at the base α. The predetermined conditions include, for example, people living in a specific place, age, gender, family composition, hobbies, and the like. For example, the server 7 sends a topic to people who live in a certain area and are participating in the mailing list before a conversation at the base α, and each respondent sends a topic from each communication terminal 9 to the server 7. Send comments on topics using the internet form.

In this case, the server 7 adds identification information to the comments collected from each respondent and stores them. The identification information is the same as that explained for the communication terminal 8 above, so the explanation will be omitted. Further, a priority order (such as a numerical value) in which comments are used in dialogue may be added to the identification information. The priority order is also the same as that explained for the communication terminal 8 above, so the explanation will be omitted. As a result, the server 7 stores the comment, identification information, and/or priority in association with each other.

[Usage image]
FIG. 2 is an image diagram of the dialogue at the base α.

As shown in FIG. 2, for example, four people participate in a meeting etc. held at the base α, and the input/output device 6 is installed on the desk 110. The input/output device 6 includes input means such as a microphone and a camera, and output means such as a microphone. The display device 2 displays materials, etc. (shared screen). The communication terminal 5 in FIG. 1 is one of the communication terminals among the participants in FIG. A display device 2 and an input/output device 6 are connected to the communication terminal 5. Further, the number of participants at the base α may be any number as long as there are two or more.

<Hardware configuration of communication terminal>
Next, the electrical hardware configuration of the communication terminal 5 will be described using FIG. 3. FIG. 3 is an electrical hardware configuration diagram of the communication terminal.

As shown in FIG. 3, the communication terminal 5 is a computer that includes a CPU 501, ROM 502, RAM 503, SSD 504, external device connection I/F (Interface) 505, network I/F 506, display 507, input device 508, and media. It includes an I/F 509 and a bus line 510.

Among these, the CPU 501 as a processor controls the operation of the communication terminal 5 as a whole. The ROM 502 stores programs used to drive the CPU 501 such as IPL. RAM 503 is used as a work area for CPU 501.

The SSD 504 reads or writes various data under the control of the CPU 501. Note that an HDD (Hard Disk Drive) may be used instead of the SSD 504.

The external device connection I/F 505 is an interface for connecting various external devices. External devices in this case include a display, speaker, keyboard, mouse, USB memory, printer, and the like.

The network I/F 506 is an interface for data communication via the communication network 100.

The display 507 is a type of display means such as liquid crystal or organic EL (Electro Luminescence) that displays various images.

The input device 508 is a keyboard, pointing device, etc., and is a type of input means for inputting, selecting, executing, etc. various instructions. Note that the input device 508 can be used in combination with an external keyboard and mouse.

The media I/F 509 controls reading or writing (storage) of data to a recording medium 509m such as a flash memory. The recording media 509m also include DVDs, Blu-ray Discs (registered trademark), and the like.

The bus line 510 is an address bus, a data bus, etc. for electrically connecting each component such as the CPU 501 shown in FIG. 4.

The communication terminal 5 may be provided with at least one of a microphone, a camera, and a speaker. Further, since the server 7 and the communication terminals 8 and 9 have basically the same hardware configuration as the communication terminal 5, the description thereof will be omitted.

Note that the communication terminal 8 at the base β is equipped with a microphone, a camera, and a speaker, and may be used instead of the above-mentioned input device and output device.

[Functional configuration of communication terminal]
The functional configuration of the communication terminal 5 according to this embodiment will be explained. FIG. 4 is a functional configuration diagram of the communication terminal 5. As shown in FIG.

As shown in FIG. 4, the communication terminal 5 includes an initial value setting section 50, a topic input section 51, a topic determination section 53, a comment selection section 55, a participant emotion determination section 57, and an output section 59. Each of these units is a function realized by instructions from the CPU 501 in FIG. 3 based on a program.

The initial value setting unit 50 accepts the input of the volume s (db) and the time T (seconds), which are used as criteria for the comment selection unit 55 to determine that silence has continued during the conversation, from participants etc. before the conversation. .

The topic input unit 51 receives as input audio data indicating the content of the conversation, including the topic being discussed. In the case of voice input using the microphone of the input/output device 6, the topic input unit 51 converts the voice input into text (character) data. The topic input unit 51 also acquires text (character) data such as comments from the communication terminal 5 or the server 7 . Text data indicating the content of the dialogue including the input topic is sent to the topic determining section 53. The text data of the content of this dialogue is also sent to the communication terminal 8 of the base β as the content of the dialogue at the base α.

The topic determination unit 53 determines the content of the topic being discussed in the dialogue. Therefore, the topic determination unit 53 morphologically analyzes the text data received from the topic input unit 51 and extracts predetermined words (for example, only nouns). The topic determining unit 53 determines the contents of frequently appearing words among the plurality of extracted words as the contents of the "topic". Whether or not a word appears frequently is determined based on whether a word appears a predetermined number of times (for example, three times) or more in a predetermined period of time (for example, 60 seconds). Note that any part of speech of the word to be extracted can be specified. In addition, the topic determination unit 53 converts the topic data in the text into higher-order coordinates using APIs such as doc2vec and fast2text, and performs principal component analysis on these coordinates. , the converted coordinates may be obtained by dimensional compression to two or three dimensions. In this case, the value of this coordinate indicates the "topic". The obtained data indicating the topic is sent to the comment selection section 55.

The comment selection unit 55 acquires the audio data of all participants present at the base α from the input/output device 6. Furthermore, the comment selection unit 55 analyzes the acquired audio data and determines whether "silence continued for a predetermined period of time" or not. For example, the comment selection unit 55 determines that "silence has continued for a predetermined period of time" when the sound volume remains below s decibels for T seconds.

Furthermore, upon acquiring topic data from the topic determining section 53, the comment selection section 55 acquires comment data related to the content of the topic from the server 7. Further, the comment selection section 55 acquires participant emotion information from the participant emotion determination section 57. Emotional information will be explained later. Then, the comment selection unit 55 selects a predetermined comment from among the comments acquired from the server 7, depending on the emotion of the participant during the dialogue indicated by the emotion information. The predetermined comment is sent to the output section 59. Note that depending on the content of the emotion, the comment selection unit 55 does not select a predetermined comment to be output, or does not send a predetermined comment that has already been selected to the output unit 59. Detailed processing of the comment selection section 55 will be explained with reference to FIG.

The participant emotion determination unit 57 determines the emotions of the participants present in the dialog. The participant emotion determination unit 57 acquires at least one of video and audio data of a plurality of participants having a dialogue at the base α from the input/output device 6. Note that instead of the microphone of the input/output device 6, a headset microphone, a lavalier microphone, a gooseneck microphone, or the like may be used to individually acquire the voices of each participant.

Further, a first example of the processing performed by the participant emotion determining unit 57 is disclosed in Reference Document 1. With the technique of Reference 1, it is possible to numerically predict the value that a person subjectively feels based on images from a camera or information on the presence or absence of speech.
<Reference 1> Junpei Otochi, Yoko Ishii, Momoko Nakatani, Kazuhiro Otsuka, "Prediction of subjective impressions of dialogue participants in multi-person dialogue using head motor functions", IEICE Technical Report, vol.121 , no.143, HCS2021-20, pp. 19-24(2021)
Further, a second example of the processing performed by the participant emotion determining unit 57 is disclosed in Reference Document 2. With the technology of Reference 2, human emotions can be predicted as numerical values based on the quality of the voice and the content of the words, using input such as uttered audio.
<Reference 2> https://group.ntt/jp/newsrelease/2021/11/01/211101b.html
The numerical values indicating the type of emotion and the degree of emotion obtained by the participant emotion judgment unit 57 are sent to the comment selection unit 55 as “emotion information” including utterance time information t when the participant (utterance) occurred. .

The output unit 59 outputs comments devised by a person from the communication terminal 8 at appropriate timings in the flow of the dialogue.

[Processing or operation of communication terminal]
Next, comment selection processing by the communication terminal 5 will be described using FIG. 5. Note that the following processing is an example of the processing of each part shown in FIG.

S11: The comment selection unit 55 acquires the audio data of all participants present at the base α from the input/output device 6.

S12: The comment selection unit 55 analyzes the acquired audio data and determines whether "silence continued for a predetermined period of time" or not. For example, the comment selection unit 55 determines that "silence has continued for a predetermined period of time" when the sound volume remains below s decibels for T seconds. If the topic determining unit 53 does not determine that silence has continued for a predetermined period of time (S12; NO), the process returns to S11. On the other hand, if the topic determining unit 53 determines that silence has continued for a predetermined period of time (S12; YES), the process proceeds to S13.

Here, in advance, the topic input unit 51 converts audio data indicating the content of the topic during the dialogue into text data, and the topic determination unit 53 converts words that have appeared a predetermined number of times or more in a predetermined time in the text data. Based on this, the content of the topic is determined and topic data is sent to the comment selection section 55.

S13: The comment selection unit 55 acquires topic data from the topic determination unit 53.

S14: The comment selection unit 55 selects a corresponding predetermined comment by searching for comment data related to predetermined identification information related to the topic among the identification information stored in the server 7. If there is one or more comments related to the predetermined identification information, the comment selection unit 55 selects one predetermined comment with the highest priority among them. If there are multiple comments with the same priority, the comment selection unit 55 randomly selects one predetermined comment from among these comments. Further, the comment selection unit 55 outputs a numerical value regarding the degree of similarity with the selected predetermined comment.

S15: The comment selection unit 55 receives emotional information (information such as type of emotion, numerical value indicating the degree of emotion, and time of utterance) of each participant in the dialogue taking place at the base α from the participant emotion determination unit 57. get.

S16: The comment selection unit 55 estimates the emotions of each participant in the dialogue based on the emotional information of each participant, as shown below.

For example, the comment selection unit 55 adds up the numerical values for each type of emotion of all the participants at the base α, and determines the emotion with the largest numerical value as the predetermined emotion for the dialog.

Alternatively, the comment selection unit 55 may divide the emotion types and numerical distribution characteristics of all participants into several categories, instead of dividing them by emotion type, and then divide the emotion of the conversation for each category into categories. You may decide. In this case, the comment selection unit 55 determines the emotion associated with the category whose characteristics of the obtained emotion type and numerical distribution are closest to each other as the predetermined emotion for the conversation.

Possible examples of emotions in a conversation include "excited," "concentrated," and "distracted."

The comment selection unit 55 controls whether or not to output comments depending on the emotion of the conversation. For example, when the type of emotion is "concentrated," the comment selection section 55 does not send data of the predetermined comment to the output section 59. This is because the participants are concentrating, so even if a predetermined comment is output, there is a possibility that the participants will not see or ignore the predetermined comment. On the other hand, if the type of emotion is "excited" or "distracted", the comment selection unit 55 sends predetermined comment data to the output unit 59.

S17: The output unit 59 outputs the comment data received from the comment selection unit 55. Note that the output unit 59 may acquire the degree of similarity (numeric value) with a predetermined comment from the comment selection unit 55, add this degree of similarity (numeric value), and output the result. As an output method, the input/output device 6 may output audio, or the display device 2 may display text information. In addition, the output content may be to output the comment as is (in the case of audio, the comment is read aloud using speech synthesis), or it may be possible to output the comment as is, such as "I think..." Predetermined sentences, words, or conjunctions may be added before and after.

Note that the process of S12 above may be performed between the process of S16 and S17. In this case, since the communication terminal 5 has already selected the predetermined comment to be output when silence continues for the predetermined period of time, the communication terminal 5 can quickly output the predetermined comment. This is effective when the communication speed in the communication network is slow.

Additionally, the process of S15 may be omitted. That is, the communication terminal 5 may output comments related to the topic without considering the atmosphere (level of emotion) of the participants at the base α.

This concludes the description of FIG. 5.

[Effects of embodiment]
As explained above, according to the present embodiment, the communication terminal 5 selects the topic of the comment made by the participant at the base α among a plurality of comments devised not by a computer but by a person (observer, respondent). Since predetermined comments related to the topic are output, it is possible to output the comments of the person who fits the topic at the appropriate timing in the flow of the dialogue.

Furthermore, since the output of a predetermined comment is controlled according to the level of emotion of the participants at the base α, it is possible to output the predetermined comment at a more appropriate timing in the flow of the dialogue.

●Supplement The present invention is not limited to the above-described embodiments, and may have the following configuration or processing (operation).

(1) The communication terminal 5 can also be realized by a computer and a program, but this program can also be recorded on a (non-temporary) recording medium or provided via the communication network 100.

(2) In the above embodiment, a notebook computer is shown as an example of the communication terminals 5, 8, and 9, but the invention is not limited to this. For example, a desktop computer, a tablet terminal, a smartphone, etc. may be used. .

(3) Each CPU 501 may be a single CPU or a plurality of CPUs.

(4) In the above embodiment, comments are stored on the server, but the present invention is not limited to this. For example, the comment may be stored in the RAM 503 or SSD 504 in the communication terminal 5. In this case, the comment selection unit 55 reads comment data stored in the communication terminal 5 and selects a predetermined comment. Note that the RAM 503 or SSD 505 in the communication terminal 5 and the server 7 are examples of storage units.

●Additional Notes The above-described embodiments can also be expressed as inventions shown below.

[Additional Note 1]
A communication terminal having a processor that outputs comments during dialogue between multiple participants,
The processor includes:
Converting audio data indicating the content of the dialogue including the topic during the dialogue into text data,
determining the content of the topic based on words that appear more than a predetermined number of times in a predetermined time in the text data;
Selecting a predetermined comment to be output by acquiring comment data related to the determined content of the topic from a storage unit that stores comment data devised by a person in advance;
outputting data of the selected predetermined comment;
communication terminal.

[Additional note 2]
The processor receives comment data related to the content of the topic from the storage unit as comment data corresponding to the identification information, based on identification information associated with each comment data and used to classify the comment content. The communication terminal according to Supplementary Note 1 to be acquired.

[Additional note 3]
The communication terminal according to supplementary note 1 or 2,
The processor includes:
Determining the emotions of the plurality of participants based on at least one of the video and audio of the plurality of participants in the dialogue place,
selecting the predetermined comment to be output according to the emotions of the plurality of participants;
communication terminal.

[Additional note 4]
The communication terminal according to supplementary note 1 or 2,
The processor includes:
Determining the emotions of the plurality of participants based on at least one of the video and audio of each of the participants in the dialogue place,
not selecting the predetermined comment to be output according to the emotions of the plurality of participants;
communication terminal.

[Additional Note 5]
The communication according to appendix 1 or 2, wherein the processor determines the content of a word that appears a predetermined number of times or more in a predetermined time from among a plurality of words extracted by morphologically analyzing the data of the text as the content of the topic. terminal.

[Additional Note 6]
3. The communication terminal according to claim 1 or 2, wherein the processor outputs data of the predetermined comment when silence continues for a predetermined period of time in the dialog.

[Additional Note 7]
A comment output method executed by a communication terminal having a processor that outputs comments during a dialogue between multiple participants, the method comprising:
The processor includes:
Converting audio data indicating the content of the dialogue including the topic during the dialogue into text data,
determining the content of the topic based on words that appear more than a predetermined number of times in a predetermined time in the text data;
Selecting a predetermined comment to be output by acquiring comment data related to the determined content of the topic from a storage unit that stores comment data devised by a person in advance;
outputting data of the selected predetermined comment;
A comment output method that does that.

[Additional Note 8]
A non-temporary recording medium on which a program for causing a computer to execute the method set forth in Supplementary Note 7 is recorded.

1 Communication system 2 Display device 5 Communication terminal 6 Input/output device 7 Server (an example of a storage unit)
8 Communication terminal 9 Communication terminal 50 Initial value setting section 51 Topic input section 53 Topic judgment section 55 Comment selection section 57 Participant emotion judgment section 59 Output section 503 RAM (an example of a storage section)
504 SSD (example of storage section)

Claims

A communication terminal that outputs comments during dialogue between multiple participants,
a topic input unit that converts audio data indicating the content of the dialogue including the topic during the dialogue into text data;
a topic determination unit that determines the content of the topic based on words that have appeared a predetermined number of times or more in the text data;
A comment that selects a predetermined comment to be output by acquiring comment data related to the content of the determined topic from a storage unit that stores comment data devised by a person in advance. a selection section;
an output unit that outputs data of the selected predetermined comment;
A communication terminal with
The comment selection unit selects comments related to the content of the topic from the storage unit as comment data corresponding to the identification information, based on identification information associated with each comment data and for classifying comment content. The communication terminal according to claim 1, which acquires data.
The communication terminal according to claim 1 or 2,
comprising a participant emotion determination unit that determines the emotions of the plurality of participants based on at least one of video and audio of the plurality of participants in the dialogue place;
The comment selection unit selects the predetermined comment to be output according to the emotions of the plurality of participants.
communication terminal.
The communication terminal according to claim 1 or 2,
a participant emotion determination unit that determines the emotions of the plurality of participants based on at least one of video and audio of the plurality of participants in the dialogue place;
The comment selection unit does not select the predetermined comment to be output according to the emotions of the plurality of participants.
communication terminal.
3. The topic determination unit determines the content of a word that appears a predetermined number of times or more in a predetermined time as the content of the topic, among a plurality of words extracted by morphologically analyzing the data of the text. communication terminal.
The communication terminal according to claim 1 or 2, wherein the output unit outputs data of the predetermined comment when silence continues for a predetermined period of time in the dialogue place.
A comment output method executed by a communication terminal that outputs comments during a dialogue between multiple participants, the method comprising:
The communication terminal is
converting audio data indicating the content of the dialogue including the topic during the dialogue into text data, determining the content of the topic based on words that appear more than a predetermined number of times in a predetermined time in the text data;
Selecting a predetermined comment to be output by acquiring comment data related to the determined content of the topic from a storage unit that stores comment data devised by a person in advance;
outputting data of the selected predetermined comment;
A comment output method that does that.
A program that causes a computer to execute the method according to claim 7.