JP6081906B2 - Discussion support device and discussion support program - Google Patents

Description

  The present invention relates to a discussion support apparatus and a discussion support program that support discussion by a plurality of participants.

  In general, in order for multiple participants to proceed appropriately and synchronously with repeated diverging and convergence, each participant must understand the current status of discussion and move toward the target discussion type. It is very important to be conscious.

  For example, when discussing while exchanging text messages on a computer, an invention for determining whether the discussion is in line with the topic by measuring the degree of coincidence between the word on the message and a predetermined topic (For example, refer to Patent Document 1). The present invention makes it possible to determine whether each speaker has a conversation according to a predetermined topic.

  In addition, an apparatus has been invented that allows a speaker to determine whether or not the explanation of conference materials is sufficient (see, for example, Patent Document 2). The invention described in Patent Document 2 automatically acquires which material the participant is viewing during the conference, so that the page viewed by the speaker and the page viewed by the participant can be obtained. Analyze the differences. The present invention can support a speaker to appropriately proceed with a conference in a conference-type dialogue in which there are a plurality of viewers for one speaker.

JP 11-39264 A JP 2000-200277 A

  However, in the above-described invention, it is considered difficult to support the progress of the discussion that a plurality of participants speaks synchronously.

  Specifically, in the invention described in Patent Document 1, it is possible to specify whether or not the discussion matches a specific topic, but specify the discussion type such as whether the discussion is diverging or converging. I can't do it. Further, in the invention described in Patent Document 2, it is assumed that the discussion proceeds according to the prepared material, and discussion in which a plurality of participants speak synchronously is not assumed. Therefore, it is difficult to support a place where everyone can discuss freely.

  Accordingly, an object of the present invention is to provide a discussion support apparatus and a discussion support program that support the progress of a discussion that a plurality of participants speaks synchronously.

  In order to solve the above problem, a first feature of the present invention relates to a discussion support apparatus that supports discussion by a plurality of participants. The discussion support device according to the first feature of the present invention includes a speech extraction unit that stores in a storage device speech time associated with a time when a participant's speech has occurred in a discussion and a word extracted from the speech. An estimator that calculates the amount of information in the discussion calculated from words and the ratio of the time at which the utterance occurs with respect to the discussion time, and estimates the estimated discussion type from the utterance ratio and information amount; An output means for outputting the estimated argument type to the output device is provided.

  Here, when a target discussion type, which is a discussion type as a discussion target, is input, it further includes a discussion target acquisition unit that stores the input target discussion type, and the output unit further displays the target discussion type as an output device. May be output.

  Further, the storage device stores message data in which a target discussion type, an estimated discussion type, and a message that supports the transition of the discussion from the estimated discussion type to the target discussion type are stored, and the output unit further includes: A message corresponding to the target discussion type stored by the discussion target acquisition unit and the estimated discussion type estimated by the estimation unit may be extracted from the message data and output to the output device.

  When the set time used for the discussion is further input, the discussion target acquisition unit stores the input set time in association with each other, and the output unit further stores the remaining set time with the elapsed time of the discussion. A message that prompts switching of the discussion type of the discussion currently in progress may be output to the output device.

  The speech extraction means further holds, for each participant, the time at which the speech occurred and the word extracted from the speech in association with each other, and the estimation means has the speech ratio of each participant with respect to the discussion time, and each participation. The discussion type is estimated for each participant from the amount of information of each participant in the discussion calculated from the person's word, and the output means further includes the target discussion type among the estimated discussion types for each participant output by the estimation means. Information on participants of different estimated discussion types may be output to the output device.

  The message of the message data includes a parameter for setting a word, and the output means sets the word extracted from the discussion utterance or the word extracted from the discussion utterance preceding the discussion. , It may be output to an output device.

  When the speech ratio is lower than the predetermined value, the estimation means stagnates the estimated discussion type, and when the speech ratio is larger than the predetermined value and the information amount is smaller than the predetermined value, the estimated discussion type converges, and the speech ratio is larger than the predetermined value. If is greater than a predetermined value, the estimated discussion type may be estimated to diverge.

  The second feature of the present invention relates to a discussion support program for causing a computer to function as the means described in the first feature of the present invention.

  ADVANTAGE OF THE INVENTION According to this invention, the discussion assistance apparatus and discussion assistance program which support the progress of the discussion which a several participant speaks synchronously can be provided.

It is a figure explaining the hardware constitutions and functional block of the discussion assistance apparatus which concern on embodiment of this invention. It is a figure explaining an example of the argument which the discussion assistance apparatus concerning embodiment of this invention makes object. It is a flowchart explaining the discussion target acquisition process in the discussion support device which concerns on embodiment of this invention. It is a figure explaining an example of the data structure and data of statement data in the discussion support device concerning an embodiment of the invention. It is a flowchart explaining the speech extraction process in the discussion assistance apparatus which concerns on embodiment of this invention. It is a flowchart explaining the estimation process in the discussion assistance apparatus which concerns on embodiment of this invention. It is a figure explaining the data structure and example of data of the message data in the discussion assistance apparatus which concerns on embodiment of this invention. It is a flowchart explaining the output process in the discussion assistance apparatus which concerns on embodiment of this invention. It is a figure explaining an example of the screen which the discussion assistance apparatus which concerns on embodiment of this invention outputs.

  Next, embodiments of the present invention will be described with reference to the drawings. In the following description of the drawings, the same or similar parts are denoted by the same or similar reference numerals.

(Embodiment)
The discussion support apparatus 1 according to the embodiment of the present invention shown in FIG. 1 supports the progress of discussions that a plurality of participants speaks synchronously. The discussion support apparatus 1 analyzes the contents of the discussions that are sequentially input, and estimates and outputs a discussion type such as whether the discussion has converged or diverged.

  In the embodiment of the present invention, the discussion support apparatus 1 describes a case in which a discussion proceeds by speaking with the voice of a participant in a conference room where a plurality of participants are present, but is not limited thereto. For example, the discussion support apparatus 1 can be applied to a case where the discussion proceeds by characters such as chat, and any device can be used as long as it is a synchronous discussion.

  The discussion targeted by the discussion support apparatus 1 proceeds by repeating divergence and convergence, for example, as shown in FIG. In the example shown in FIG. 2, each discussion section K1, K2, K3, and K4 and set times T1, T2, T3, and T4 of each discussion section are set in advance by participants, organizers, and the like. In the discussion section, “divergence” or “convergence” is set as the target discussion type. In the example illustrated in FIG. 2, the case where divergence and convergence are alternately repeated will be described. However, the discussion section of divergence may be set continuously, and is not limited to the example illustrated in FIG. 2.

  With reference to FIG. 1, the discussion support apparatus 1 which concerns on embodiment of this invention is demonstrated. The discussion support device 1 is a general computer including a storage device 10, a central processing control device 20, an input device 30, an output device 40, and the like. The storage device 10 stores data that the central processing control device 20 refers to or updates. The input device 30 is, for example, a keyboard or a microphone. The input device 30 inputs, to the central processing control device 20, instructions from participants and organizers of the discussion and voice data of the comments in the discussion. The output device 40 is a display, for example. The output device 40 outputs the output of the central processing control device 20 so that participants of the discussion can recognize.

  The storage device 10 stores the discussion support program according to the embodiment of the present invention, and also stores target discussion type data 11, utterance data 12, estimated discussion type data 13, and message data 14. By executing the discussion support program, the central processing control device 20 functions as a discussion target acquisition unit 21, a speech extraction unit 22, an estimation unit 23, and an output unit 24.

  The discussion target acquisition unit 21 causes the input device 30 to input, for a predetermined discussion section, participants, organizers, and the like, a target discussion type that is a discussion type targeted for discussion, and a set time used for this discussion. When the target discussion type and the set time are input for a predetermined discussion section, the discussion target acquisition unit 21 stores the target discussion type data 11 in which the input target discussion type and the set time are associated with each other in the storage device 10. Remember.

  Here, the input target argument type is, for example, “divergence” or “convergence”. The discussion target acquisition means 21 may display buttons corresponding to “divergence” and “convergence” on the output device 40 and allow the participants to select with the input device 30, or allow the participants to “diverge”. Alternatively, “convergence” may be input by voice or a keyboard. Any method may be used as long as the target discussion type selected by the participants can be acquired as data that can be identified by the discussion target acquisition unit 21.

  For example, when a discussion that diverges for 15 minutes proceeds, the discussion target acquisition means 21 associates the target discussion type “divergence” with the set time “15 minutes” and stores them in the target discussion type data 11. As shown in FIG. 2, when one discussion is executed through a plurality of discussion sections, the discussion target acquisition means 21 associates a target discussion type and a set time with each discussion section, and sets target discussion type data. 11 may be stored.

  With reference to FIG. 3, the discussion target acquisition process by the discussion target acquisition means 21 is demonstrated. First, in step S101, the participants wait for input of a target discussion type. When the target discussion type is input, the discussion target acquisition unit 21 stores the input discussion type as the target discussion type data 11 in step S102. Here, when a time for achieving the target discussion type is input, the target discussion type data 11 is stored by associating the input time with the input discussion type.

  The speech extraction means 22 sequentially analyzes the speech data of the discussion input from the input device 30, and uses the speech data 12 associated with the time when the participant's speech occurred in the discussion and the word extracted from the speech. Store in the storage device 10. Here, the utterance extraction unit 22 may hold the time when the utterance occurs and the word extracted from the utterance in association with each participant. The speech extraction means 22 updates the speech data 12 sequentially during the discussion, and causes the estimation means 23 described later to estimate the current discussion type based on the current speech.

  As shown in FIG. 4, the utterance data 12 is data in which a time when a participant makes a utterance, a speaker, a utterance time, and a word group extracted from the utterance. The word group extracted here includes part-of-speech words such as nouns and verbs that are easy to estimate the content of a statement, and may not include particles or auxiliary verbs.

  In the embodiment of the present invention, the speech extraction unit 22 is configured to input, for a section in which voice of the voice data is input at a sound volume equal to or higher than a certain threshold among the input audio data of the discussion. Considered a remark. Here, the volume and time thresholds are set as appropriate. For example, in a state where the discussion support device 1 is placed at the discussion holding place, first, the non-speaking state is calibrated without speaking, and then the participants speak for a certain period of time, and the average value of the volume is taken. Can be determined. Further, the threshold values of the volume and time may be manually input, may be determined separately from data recorded on the state of discussion, and may be determined by any means.

  The utterance extraction unit 22 extracts words from the data extracted as utterances. Regarding word extraction, for example, after speech is converted to text using speech recognition technology (for example, see http://www.ntt.co.jp/svlab/activity/category_2/product2_12.html) Extraction can be realized by performing morphological analysis on the text data. Furthermore, the utterance extraction unit 22 may specify the speaker of each utterance based on voice quality, voiceprint, or the like, or may identify each speaker by attaching a microphone to the utterance.

  With reference to FIG. 5, the speech extraction process by the speech extraction means 22 is demonstrated. First, in step S201, the message extraction unit 22 determines whether there is a voice input. Specifically, when the volume is higher than a predetermined level, it is determined that there is a voice input. If there is a voice input, it is determined in step S202 whether or not it can be determined as a speech. Specifically, if this predetermined or higher volume continues for a predetermined time or more, it is determined that the sound is generated.

  If it is determined as pronunciation, the speech extraction means 22 converts the speech data into text in step S203 and extracts words from the speech text in step S204. Furthermore, in step S205, the word group extracted in step S204 and the duration time of the speech regarded as a speech are stored in the speech data 12. At this time, the utterance extraction unit 22 may store the information of the speaker specified from the utterance in the utterance data 12 in association with each other.

  In step S206, the speech extraction unit 22 sequentially repeats the processing from step S201 to step S205 until the discussion ends, and the speech extraction processing ends as the discussion ends.

  The estimation means 23 refers to the utterance data 12 and calculates the utterance ratio, which is the ratio of the time when the utterance occurred with respect to the discussion time, and the information amount in the discussion calculated from the words, and from the utterance ratio and the information amount, Estimate the estimated discussion type. When the speech ratio is lower than the predetermined value, the estimating means 23 sets the estimated discussion type to “stagnation”, and when the speech ratio is larger than the predetermined value and the information amount is smaller than the predetermined value, the estimated discussion type is “convergence” and the speech ratio is predetermined. If the information amount is greater than the value and greater than the predetermined value, the estimated discussion type is estimated to be “divergence”.

  Here, the estimation means 23 may estimate the discussion type for each participant. Specifically, the estimation means 23 estimates the discussion type for each participant from the speech ratio of each participant with respect to the discussion time and the information amount of each participant in the discussion calculated from the words of each participant. The estimation means 23 stores the estimated discussion type and the discussion type of each participant in the storage device 10 as the estimated discussion type data 13.

  The estimation means 23 reads the utterance data 12 at regular intervals, for example, every 5 minutes, and estimates the current discussion type. The utterance ratio and the information amount are calculated for a word group extracted in the discussion section that is currently in progress. For example, the estimation unit 23 may estimate the discussion type based on the utterance data 12 after the discussion section of the target discussion type of “divergence” starts, or the utterance data 12 of a predetermined time such as the latest 15 minutes. The discussion type may be estimated based on the above.

  First, the estimating means 23 calculates a speech ratio. The utterance ratio is an index indicating whether the amount of utterance is large or small, and is the ratio of the time when the utterance occurs to the discussion time. The speech ratio is expressed as a ratio of the speech time in the discussion elapsed time, and 0 ≦ the speech ratio ≦ 1.0. A large or small amount of speech may be determined based on an average speech rate. The average utterance ratio may be set based on the average of the utterances of the discussion participants before starting the discussion, or may be set manually.

Next, the estimation means 23 reads the utterance data 12 and calculates the amount of information based on each word group. The information amount E can be calculated based on the equation (1).
[Equation 1]
Probability that word appears in all words P (word [i]) = number of occurrences of word [i] / number of occurrences of all words E (word group) = Σ (P (word [i]) x log P (word [i]))
(Natural number of i = 1 to n) (1)
When the utterance ratio and the information amount are calculated, the estimation unit 23 estimates the discussion type based on the calculated utterance ratio and information amount. Specifically, (a) “divergence” when the statement ratio is high and the information amount is large, (b) “convergence” when the statement ratio is high and the information amount is small, and (c) when the statement ratio is low. Estimates the type of discussion as “stagnation”.

Here, the threshold value for determining the amount of information is determined by an arbitrary method. For example, the participants are allowed to make a discussion that diverges with a large amount of speech and a large amount of information and a discussion that converges with a large amount of speech and a small amount of information in advance (m times). Based on the utterances and word groups in those discussions, the average and variance of the information amount in divergence and convergence are calculated using them as learning data. Furthermore, it is possible to set a threshold value for determining the amount of information based on the calculated average and variance.
[Equation 2]
Divergence average Avg_D = Avg (E (Divergent word group [j]))
Convergence average Avg_C = Avg (E (Convergence learning word group [j]))
(Avg is a function for obtaining an average, j = 1 to m is a natural number) Expression (2)
Here, in divergent or convergent learning data, the amount of information may be calculated using a word group extracted at an arbitrary time, and the passage of time and a change in the amount of information may be obtained. The estimation means 23 may estimate the discussion type according to the passage of time, based on the index of change in time and information amount in divergence and the indicator of change in time and information amount in convergence.

  With reference to FIG. 7, the estimation process by the estimation means 23 is demonstrated. First, in step S301, the estimation unit 23 reads the utterance data 12 from the storage device 10 and calculates the utterance ratio. In step S302, it is determined whether the speech ratio is equal to or greater than a predetermined value. If it is not equal to or greater than the predetermined value, in step S306, the estimation unit 23 determines that the estimated discussion type is “stagnation”.

  On the other hand, if it is greater than or equal to the predetermined value, the estimating means 23 extracts a word group from the utterance data 12 in step S303, and calculates the information amount of the word group extracted in step S304. In step S305, the estimation unit 23 determines whether or not the information amount is equal to or greater than a predetermined value, specifically, a threshold value for determining the amount of information. If it is greater than or equal to the predetermined value, the estimation means 23 determines that the estimated discussion type is “divergence” in step S307. If it is not equal to or greater than the predetermined value, the estimation means 23 determines that the estimated discussion type is “convergence” in step S308.

  In step S309, the estimation unit 23 stores the estimated discussion type determined in steps S306 to S307 as the estimated discussion type data 13. Here, when estimating the type of discussion for each participant, in step S301 and step S303, the utterance data of each participant is filtered, and the utterance ratio and information amount are calculated for each participant. Determine the discussion type.

  The output unit 24 outputs the estimated discussion type estimated by the estimation unit 23 to the output device 40 so that participants of the discussion can recognize. The output means 24 may display the estimated discussion type on a display, or may convert it into a voice message and output it from a speaker. Here, the output unit 24 further outputs the target discussion type to the output device 40. By outputting both the estimated discussion type estimated by the estimation means 23 and the target discussion type input by the participants, it becomes easier for the participants to recognize the difference between the target and the current state.

  Here, the output unit 24 further extracts, from the message data 14, a message corresponding to the target discussion type stored in the discussion target acquisition unit 21 and the estimated discussion type estimated by the estimation unit 23, and outputs it to the output device 40. It may be output. The message data 14 is data in which a target discussion type, an estimated discussion type, and a message that supports the transition of the discussion from the estimated discussion type to the target discussion type are associated with each other.

  For example, as shown in FIG. 7, the message data 14 is data in which an ID for identifying a record, a target argument type, an estimated argument type, a remaining time, a category, and a message are associated with each other. The message shown in FIG. 7 is an example, and the message is not limited to this. Any message may be used as long as it can be derived from the current state to the target discussion type.

  The category is associated with the target argument type, the estimated argument type, and the remaining time, and classifies messages. The output unit 24 outputs a message prompting the switching of the discussion type of the discussion currently in progress to the output device 40 when the remaining set time is small with the elapsed discussion time. As a result, the participants can be encouraged to improve the outcome of the discussion within the design time given in the discussion section.

  In the example illustrated in FIG. 7, any of diverging advice, convergence advice, and switching advice is set as the category. "Diverging advice" is a message type for encouraging participants to lead discussions to a diverging tendency. “Convergence advice” is a message type for encouraging participants to guide discussions toward convergence. “Switching advice” is a message type for encouraging participants to guide discussions to a different trend from the current time, or to guide them to a target discussion type when there is little remaining time.

  The remaining time is the remaining time from the timing of issuing a message to the end of the discussion section in the discussion section that is progressing when the message is output. The remaining time “large” indicates a case in which a message is output after a predetermined time elapses after discussion in a predetermined discussion section starts. The remaining time “less” indicates a case where a message is output after a predetermined time elapses in the discussion of the predetermined discussion section and before the set time of the predetermined discussion section expires. The threshold value of the time for determining “more” or “less” is determined in advance. For example, when the set time of the discussion section is set to 30 minutes and the estimation means 23 estimates the estimated discussion type every 5 minutes, the timing when 5 minutes, 10 minutes, and 15 minutes have elapsed from the start is expressed as “ "There are many". On the other hand, at the timing when 20 minutes and 25 minutes have elapsed from the start, the remaining time is “less”. In the example shown in FIG. 7, “remaining time” is classified into “high” and “low”, but is not limited thereto. For example, the remaining time in the discussion section may be set as “remaining time”, and the message may be set according to each classification.

  Since the output means 24 can output the message in consideration of the remaining time by the message data 14, the time set for each discussion section is used effectively, and depending on the timing of outputting the message, it is appropriate. Message can be output.

  Further, the message of the message data 14 may include a parameter for setting a word. In this case, the output unit 24 sets a word extracted from the utterance utterance as a parameter and outputs it to the output device 40. The output unit 24 sets a word extracted from the utterance of the discussion performed prior to the discussion as a parameter, and outputs the word to the output device 40.

  As shown in FIG. 2, it is assumed that the discussion according to the embodiment of the present invention is performed through a plurality of discussion sections. Therefore, by setting a word that appeared in the discussion section that is currently in progress or a word that appeared in the discussion section that ended before in the parameter, a specific message is output to the participants, and appropriate discussion progress Can help.

  This word is a word extracted by the speech extraction means 22. Specifically, when a parameter is included in the message output in the discussion section K2, words extracted so far in the discussion section K2, words extracted in the discussion section K1, and the like are set. In addition, the parameters may be set with words that can be identified from the utterance data 12, such as words that appear most frequently and words that appear from each participant. What word is set is specified by including a code indicating the type in the parameter of the message data.

  Specifically, in the case of the category “divergence advice”, the output unit 24 searches for words extracted in the discussion so far from the utterance data 12 and sets the searched words as parameters. Further, in the case of the category “convergence advice”, the output means 24 is a word extracted from the speech data 12 in the discussion section that ended before the discussion section currently in progress, but is extracted in the discussion section currently in progress. Search for a word that is not, and set the searched word as a parameter.

  The output unit 24 may change the information to be output according to the elapsed time of the discussion section currently in progress. For example, as the set time of the discussion section that is currently in progress approaches, the message or the estimated discussion type may be displayed or the message may be displayed more frequently.

  The output unit 24 may output the estimated discussion type for each participant to the output device 40 when the estimation unit 23 outputs the estimated discussion type for each participant. At this time, the output unit 24 may identify a participant having an estimated discussion type different from the target discussion type and output the specified participant to the output device 40. For example, the message data 14 associates a message including a parameter for specifying a participant name with a target discussion type and an estimated discussion type for each participant. Specifically, if the target discussion type is “divergence” and Mr. A's estimated discussion type is “stagnation”, the output means 24 will say “What do you think about Mr. A?” It is possible to output a prompt message. Moreover, you may notify that the discussion is progressing by the presumed argument classification different from a target argument classification so that it may be transmitted only to the identified participant.

  With reference to FIG. 8, the output process by the output unit 24 according to the embodiment of the present invention will be described.

  In step S <b> 401, the output unit 24 reads the message data 14, acquires the target discussion type from the target discussion type data 11, and acquires the estimated discussion type from the estimated discussion type data 13. In step S403, the output unit 24 acquires a message corresponding to the target argument type and the estimated argument type acquired in step S402 from the message data 14. At this time, the remaining time may be calculated as necessary, and the message may be acquired in consideration of the remaining time. In step S404, the output unit 24 outputs the message acquired in step S403 to the output device 40.

  With reference to FIG. 9, the screen which the output means 24 which concerns on embodiment of this invention displays on the output device 40 is demonstrated. Each screen P includes a target discussion type display unit D1, an estimated discussion type display unit D2, a message display unit D3, and a remaining time display unit D4.

  FIG. 9A is a screen P1 displayed when the target discussion type is “divergence” and the estimated discussion type is “divergence”. The target discussion type display unit D1 displays the area described as “diverging” more prominently than the area described as “convergence”. The estimated discussion type display unit D2 displays “divergence tendency”. Here, when there is much remaining time, the message display part D3 displays the message which further promotes divergence.

  FIG. 9B is a screen P2 displayed when the target discussion type is “divergence” and the estimated discussion type is “convergence”. The target discussion type display unit D1 displays the area described as “diverging” more prominently than the area described as “convergence”. The estimated discussion type display unit D2 displays “convergence tendency”. Here, when the remaining time is short, the message display unit D3 displays a message for promoting convergence. Further, the message display unit D3 also displays buttons B1 and B2 for selecting whether to continue discussion of the divergence tendency or to converge. When the button B1 for continuing the discussion is selected, the set time set in the discussion section is extended and the display of the remaining time display section D4 is updated. Further, the output unit 24 acquires a message from the message data 14 based on the extended set time and displays it on the message display unit D3. In the example shown in FIG. 9B, the converging button B2 is displayed larger than the continuing button B1 in order to respect the set time set in advance in the discussion section.

  FIG. 9C is a screen P3 displayed when the target discussion type is “convergence” and the estimated discussion type is “divergence”. The target discussion type display unit D1 displays the area described as “converging” more prominently than the area described as “divergence”. The estimated discussion type display unit D2 displays “divergence tendency”. Here, when the remaining time is large, the message display unit D3 displays a message for prompting convergence.

  When the discussion support apparatus 1 according to the embodiment of the present invention discusses while repeating divergence and convergence, each participant can clearly check the current discussion type. Furthermore, the discussion support apparatus 1 can easily determine whether or not the discussion is appropriately performed for the target discussion type based on the speech state of the participant. If the discussion support apparatus 1 is not proceeding properly, the discussion support apparatus 1 immediately outputs a message indicating that the discussion type should be changed, and makes it easier for the participants to determine the next discussion type to be selected. Can be encouraged.

  As described above, the discussion support apparatus 1 according to the embodiment of the present invention can smoothly and appropriately support the progress of the discussion that a plurality of participants speaks synchronously.

(Other embodiments)
As described above, the embodiments of the present invention have been described. However, it should not be understood that the descriptions and drawings constituting a part of this disclosure limit the present invention. From this disclosure, various alternative embodiments, examples, and operational techniques will be apparent to those skilled in the art.

  For example, the discussion support apparatus described in the embodiment of the present invention may be configured on a single piece of hardware as shown in FIG. 1, or may be configured on a plurality of pieces of hardware according to the functions and the number of processes. May be. Further, the discussion support device may be configured by a client terminal installed in a discussion venue where discussions proceed and a connection server device connected to the client terminal via a communication network.

  It goes without saying that the present invention includes various embodiments not described herein. Therefore, the technical scope of the present invention is defined only by the invention specifying matters according to the scope of claims reasonable from the above description.

DESCRIPTION OF SYMBOLS 1 Discussion support apparatus 10 Memory | storage device 11 Target discussion classification data 12 Statement data 13 Estimated discussion classification data 14 Message data 20 Central processing control device 21 Discussion target acquisition means 22 Statement extraction means 23 Estimation means 24 Output means 30 Input apparatus

Claims (8)

A discussion support device that supports discussion by a plurality of participants,
A speech extraction means for storing in a storage device the speech occurrence time of the participant's speech in the discussion and the speech data associated with the word extracted from the speech;
A speech ratio that is a ratio of the time at which the speech occurs with respect to the time of the discussion; ,
An argument support device comprising: output means for outputting the estimated discussion type estimated by the estimation means to an output device. When a target discussion type, which is a discussion type that is a target of the discussion, is input, further comprising a discussion target acquisition unit that stores the input target discussion type,
The discussion support apparatus according to claim 1, wherein the output unit further outputs the target discussion type to the output device. The storage device stores message data associating a target discussion type, an estimated discussion type, and a message for supporting transition of the discussion from the estimated discussion type to the target discussion type,
The output means further extracts, from the message data, a message corresponding to the target discussion type stored by the discussion target acquisition means and the estimated discussion type estimated by the estimation means, and outputs the message to the output device. The discussion support apparatus according to claim 2 , wherein the apparatus is a discussion support apparatus. When the set time used for the discussion is input, the discussion target acquisition means further stores the input set time in association with each other,
The output means further outputs, to the output device, a message prompting switching of the discussion type of the discussion currently in progress when the remaining set time is small with the elapsed time of the discussion. Item 4. The discussion support apparatus according to Item 2 or 3. The speech extraction means further holds, for each participant, the time when the speech is generated and the word extracted from the speech in association with each other,
The estimation means estimates the discussion type for each participant from the speech ratio of each participant with respect to the discussion time and the information amount of each participant in the discussion calculated from the word of each participant,
The output means further outputs, to the output device, information on participants of an estimated discussion type different from the target discussion type among the estimated discussion types for each participant output by the estimation means. The discussion support device according to any one of claims 2 to 4. The message data message includes a parameter for setting a word,
The output means sets, in the parameter, a word extracted from the discussion utterance or a word extracted from the discussion utterance that preceded the discussion, and outputs it to the output device. The discussion support apparatus according to claim 3. When the speech ratio is lower than a predetermined value, the estimation means stagnates the estimated discussion type. The discussion support apparatus according to any one of claims 1 to 6, wherein when the amount is greater than a predetermined value, the estimated discussion type is estimated to diverge. 8. A discussion support program for causing a computer to function as the discussion support device according to claim 1.

Images