JP2010175778A - Interview support method, device and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide technology for supporting interview up to speed. <P>SOLUTION: When a response interval or an speech speed of an interviewee is different from usual, the interviewee may not respond in a good pace in the interview, so that an interviewer is prompted to adjust his or her speech intervals. Thereby, suitable response is expected from the interviewee. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  This technology relates to interview support technology.

  In response to the voice message from the user, the user responds to the sent response message in an interactive voice response device such as an answering machine. Use the threshold that determines the absence of willingness to start (silent state determination threshold) and the threshold that determines the end of the utterance of the user who started speaking (utterance end determination threshold) that is optimal for response message response Therefore, there is a technology that allows the dialogue with the user to proceed more smoothly. A threshold corresponding to the content of the response message is set. Such a technique cannot set an appropriate threshold unless the content of the response message can be understood.

Japanese Patent No. 2590193

  When conducting an interview rather than in a field such as an interactive voice response device, the answer may be interrupted, such as asking the next question before the respondent of the interview answers. In addition, even if the respondent has finished answering, the answerer may be dismissed without the questioner immediately answering the next question. If questions and answers are not made at such a fast pace, the respondents' thoughts may go in a direction different from the intent of the questioner or spread, and appropriate answers cannot be drawn from the respondent. There is a case. On the other hand, in interviews, questions can be prepared in advance to some extent, but the contents of responses from respondents cannot be prepared in advance, and it is necessary to plan an appropriate timing each time.

  Therefore, an object of the present technology is to provide a technology for assisting in conducting an interview at an appropriate tempo.

  This interview support method is based on the current period of the respondent, or the current speaking speed of the respondent, which is the time from the end of the commenter's speech to the start of the respondent's speech within the predetermined time during the first interview. From the first interview after the acquisition step of acquiring the current time of the questioner, which is the time from the respondent's speech end to the questioner's speech end, and after detecting the questioner's speech end during the first interview Answers from the reference data storage unit that stores the reference interval between respondents or the reference rate of respondents, which is the reference time from the end of the interviewer's speech to the start of the respondent's speech, collected in the previous second interview Steps for reading between the respondent's criteria or the respondent's reference speech rate and comparing the respondent's current speech rate with the respondent's criteria, or comparing the respondent's current speech rate with the respondent's reference speech rate And a condition that the first predetermined condition including a condition that the current time of the respondent and the standard of the respondent do not match is satisfied, or that the current speaking speed of the respondent and the reference speaking speed of the respondent do not match. When the second predetermined condition is included, the current time of the questioner, the relationship between the current time of the respondent and the standard of the respondent, or the relationship between the current speech rate of the respondent and the reference speech rate of the respondent And a step of presenting to the questioner the current state of the questioner after the change when the end of the speech of the respondent is detected during the first interview.

FIG. 1 is a functional block diagram of the interview support apparatus. FIG. 2 is a diagram for explaining parameters in the present embodiment. FIG. 3 is a diagram showing a main process flow of a phase process for making a basic question. FIG. 4 is a diagram illustrating an example of the initial screen. FIG. 5 is a diagram illustrating an example of an interview log basic data table. FIG. 6 is a diagram illustrating an example of an interview state transition. FIG. 7 is a diagram illustrating a processing flow of parameter calculation processing. FIG. 8 is a diagram for explaining the process of changing the interview log basic data table. FIG. 9 is a diagram illustrating an example of the processed interview log basic data table. FIG. 10 is a diagram illustrating a process flow of the parameter calculation process. FIG. 11 is a diagram illustrating an example of the second screen. FIG. 12 is a diagram illustrating an example of the state transition of the interview in the phase for asking the question of the main subject. FIG. 13 is a diagram illustrating an example of the third screen. FIG. 14 is a diagram illustrating a main processing flow of the phase processing for asking the question of the main subject. FIG. 15 is a diagram illustrating an example of an interview log data table. FIG. 16 is a diagram illustrating a processing flow of calculation processing between the optimum. FIG. 17A is a diagram illustrating an example of the interview log data table after the change. FIG. 17B is a diagram for explaining a range of processing target records. FIG. 18 is a diagram illustrating a process flow of the inter-adjustment process. FIG. 19 is a diagram illustrating a process flow of the inter-adjustment process. FIG. 20 is a diagram showing a processing flow of display processing between the optimum. FIG. 21 is a functional block diagram of a computer.

  FIG. 1 shows a functional block diagram of an interview support apparatus according to an embodiment of the present technology. The interview support device 5 is connected to, for example, two microphones 1 and 3 and can capture the voices of the respondent and the questioner. The interview support apparatus 5 includes an audio-related data processing unit 51 that processes audio data from the microphones 1 and 3, a log data storage unit 53 that stores data processed by the audio-related data processing unit 51, and an audio-related data. A control unit 55 that performs processing using data stored in the log data storage unit 53 in cooperation with the data processing unit 51; a parameter storage unit 57 that stores values of various parameters calculated by the control unit 55; And a display device 58.

  In addition, the control unit 55 includes a basic data processing unit 551, an optimal interval calculation unit 553, a parameter calculation unit 552, and an input / output unit 554.

  Next, the processing contents of the interview support apparatus 5 will be described separately in (1) phase processing for performing a basic question and (2) phase processing for performing a main question.

1. Processing of the phase for asking basic questions In this phase, the questioner asks questions that are easy for the respondents to answer, such as the respondent's name, affiliation, work history, and business details, and obtains the parameters described below as basic data. To do.

First, parameters required in the present embodiment will be described with reference to FIG. In FIG. 2, time is assumed to flow from left to right. In the interview, the question from the questioner and the answer from the respondent are repeated. Here, the time from the end of the questioner's speech to the start of the respondent's speech is referred to as between the respondents. The time from the end of the respondent's speech to the start of the questioner's speech is referred to as between the questioners. In FIG. 2, there are two times between questioners (that is, M ₁ and M ₂ ), and M between questioners used as parameters below is the average value ((M ₁ + M ₂ ) / 2). It is. Similarly, two times occur between respondents (ie, T ₁ and T ₂ ), and T among respondents used as a parameter is an average value ((T ₁ + T ₂ ) / 2). is there. Further, in each phase, the total question time Tq (= T _q1 + T _q2 + T _q3 ) and the total answer time Ta (= T _a1 + T _a2 ) are also calculated.

Further, in the present embodiment, the question time and the answer time are recognized by speech recognition and the number of words is counted, and the questioner's speech rate A is calculated by the total number of questions Tword / question time T _q . and, the speech rate B of the respondents, calculates the sum T _a number of words / answer time answers.

  It should be noted that M between the interrogators calculated in this phase is denoted as Basic_M, T between the respondents is denoted as Basic_T, the speaking rate A of the questioner is denoted as Basic_A, and the speaking rate B of the respondent is denoted as Basic_B.

  Next, processing in a phase for performing a basic question will be described with reference to FIGS. 3 to 11. For example, in response to an instruction from the questioner, the input / output unit 554 of the control unit 55 displays an initial screen on the display device 58 (step S1). For example, a screen as shown in FIG. 4 is displayed. As shown in FIG. 4, a start button and an end button for instructing start and end of data collection in this phase, and a table for displaying the speaking rate and interval between the questioner and the answerer are displayed. However, numerical values are not yet displayed for speaking rate and interval.

  Next, when the questioner clicks a start button to start the question, the input / output unit 554 receives the start instruction and notifies the basic data processing unit 551 of the start instruction. On the other hand, the basic data processing unit 551 instructs the audio related data processing unit 51 to start processing, and the audio processing related data processing unit 51 records audio data from the microphones 1 and 3 (steps). S2). For example, the digitized voice data is stored in the log data storage unit 53 as it is. Further, the voice related data processing unit 51 generates an interview log basic data table from the voice data and stores it in the log data storage unit 53 (step S3). The interview log basic data table is, for example, a table as shown in FIG. In FIG. 5, the start time and end time of an utterance (also referred to as an utterance), the speaker, and the identifier of voice data from the start time to the end time are registered.

  The data registered in the interview log basic data table of FIG. 5 indicates the state of conversation as shown in FIG. That is, the utterance 1 only for the questioner corresponds to the first record of the interview log basic data table, the silent section 2 corresponds to the second record, the utterance 3 only for the respondent corresponds to the third record, and the silent section 4 corresponds to the fourth record, the utterance 5 of the respondent only corresponds to the fifth record, the silent section 6 corresponds to the sixth record, the utterance 7 of the respondent only corresponds to the seventh record, and the silent section 8 corresponds to the 8th record, the utterance 9 of the questioner only corresponds to the 9th record, the silent section 10 corresponds to the 10th record, and the time when the questioner and the respondent speak simultaneously and compete 11 Corresponds to the eleventh record.

  As shown in the first record of FIG. 5, the voice-related data processing unit 51 records the start time when detecting the utterance of the interrogator from the microphone 1, for example, and then records the end time when detecting the silent section. The data indicating the questioner and the identifier of the voice data from the start time to the end time are recorded. The same process is performed even if the respondent is not the questioner. However, “respondent” is registered in the speaker column. Furthermore, the time from the start time to the end time is also recorded. The time is recorded in the same way regardless of the utterance of the questioner, the utterance of the respondent, the silent section, or the competition.

  Further, as shown in the second record of FIG. 5, when a silent section is detected, a start time is recorded, and when any speaker's utterance is detected, an end time is recorded, and data representing silence is provided to the speaker. Record. Furthermore, when it was detected that both the respondent and the questioner were speaking, the start time was recorded as shown in the eleventh record in FIG. 5, and then either the speaker became silent or one of the speakers stopped speaking. Is detected, the end time is recorded, the data representing the competition as the speaker is recorded, and the identifiers of the voice data of the respondent and the questioner are recorded.

  Furthermore, the voice related data processing unit 51 performs a known voice recognition process on the voice data of each record, identifies words, counts the number of words, and registers them in the corresponding records.

  Such processing is continuously performed until an end is instructed.

  Thereafter, when the end button is clicked by the questioner, the input / output unit 554 receives the end instruction and outputs the end instruction to the basic data processing unit 551. On the other hand, the basic data processing unit 551 instructs the parameter calculation unit 552 to perform processing, and the parameter calculation unit 552 performs parameter processing on the interview log basic data table stored in the log data storage unit 53 according to the instruction. Calculation processing is performed (step S5). The parameter calculation process will be described with reference to FIGS.

  First, the parameter calculation unit 552 specifies one unprocessed silent section from the top of the table to be processed (step S11). And it is confirmed whether the record exists before and behind the record of the specified silence area (step S13). The case where there is no record before and after the record of the specified silent section is either the first record or the last record of the processing target table. If there is no record before and after the record of the specified silent section, the processing target is set to the speaker column for the specified record (step S15). The first and last silent sections are not considered below. Then, control goes to a step S23.

  On the other hand, when there are records before and after the record of the specified silent section, the parameter calculation unit 552 determines whether the speakers of the records before and after the record of the specified silent section match ( Step S17). A part of the interview log basic data table is shown on the left side of FIG. 8. For example, when the second record is a silent section record to be processed, the speakers of the preceding and following records are different. On the other hand, if the fourth record is a silent section record to be processed, the speakers in the preceding and succeeding records match. If the speakers of the record before and after the specified silent section record do not match, the “speaker name of the record immediately after” of the specified silent section record is It is set as a record speaker (step S21). In the case of the second record of FIG. 8, as shown in the table on the right side of FIG. 8, between the speaker names of the next record, that is, “between the respondents” is registered. In the table on the right side of FIG. 8, it is registered in the column of the speaker (after processing). Then, control goes to a step S23.

  On the other hand, when the speakers of the record before and after the record of the specified silence section match, the parameter calculation unit 552 sends the speaker name of the record before and after the speaker of the record of the specified silence section. Is set (step S19). In the case of the fourth record in FIG. 8, the speaker names of the preceding and succeeding records, that is, “respondent” are registered. Then, control goes to a step S23.

  After step S15, S19, or S21, the parameter calculation unit 552 determines whether all silent sections have been processed (step S23). If there is an unprocessed silent section, the process returns to step S11.

  If all silent sections are processed for the interview log basic data table of FIG. 5, an interview log basic data table as shown in FIG. 9 is obtained. For the fourth and sixth records, the speaker name is changed to the respondent, the speaker name of the eighth record is changed to “between the questioners”, and the speaker name of the tenth record is “ It has been changed to “During Conflict”.

On the other hand, when processing is performed for all silent sections, the parameter calculation unit 552 calculates the total time T _{q of the} question time by adding the time of the record in which the speaker is the questioner from the processing target table. Store in the storage unit 57 (step S25). In the example of FIG. 9, 30 + 20 = 50 seconds is calculated as T _q .

Furthermore, parameter calculation unit 552 calculates the sum T _a answer time by adding the time of the speaker is respondent record from the processing target table, stored in the parameter storage section 57 (step S27). In the example of FIG. 9, 10 + 5 + 10 + 5 + 10 = 40 seconds is calculated as T _a. The processing shifts to the processing in FIG.

  Note that, for example, the audio data of “Yes” or “I see” is registered in advance in the audio-related data processing unit 51 or the like, and the record corresponding to only the relevant audio is deleted or recognized as the audio The phrase may be excluded from the number of words, or the time of the relevant question may be deleted.

  Shifting to the description of the processing in FIG. 10, the parameter calculation unit 552 adds the time of the record in which the speaker is “between the questioners” from the table to be processed, and divides it by the number of the corresponding records. M is calculated and stored in the parameter storage unit 57 (step S29). For example, in the example of FIG. 9, since there is only one record for “between the questioners”, the time of 10 seconds is set as M between the questioners.

  Further, the parameter calculation unit 552 calculates the inter-responder T by adding the time of the record in which the speaker is “between the respondents” from the table to be processed and dividing by the number of the corresponding records. Store in the storage unit 57 (step S33). For example, in the example of FIG. 9, “between respondents” has only one record, and thus 10 seconds, which is the time, is set as T between respondents.

The parameter calculation unit 552, a speaker from the processed table to calculate the number of words of the sum of the records is a questioner, by dividing the sum T _q question time, the speech speed A of the questioner Calculate and store in the parameter storage unit 57 (step S35). In the example of FIG. 9, A = (20/50) = 0.4 words / second is calculated.

The parameter calculation unit 552, a speaker from the processed table to calculate the number of words of the sum of the records is a respondent, by dividing the sum T _a of reply time, the speech speed B of respondents Calculate and store in the parameter storage unit 57 (step S37). In the example of FIG. 9, B = (15/40) = 0.375 words / second is calculated.

  By performing the processing as described above, the parameters for the processing target table are calculated. If the processing target table is an interview log basic data table, basic data is generated. That is, Basic_M between the interrogators, Basic_T between the respondents, the speaking rate Basic_A of the interrogator, and the speaking rate Basic_B of the respondent are obtained.

  Returning to the description of the processing in FIG. 3, upon receiving a processing end notification from the parameter calculation unit 552, the basic data processing unit 551 receives the questioner, which is basic data stored in the parameter storage unit 57, in the input / output unit 554. Then, the speaking rate and interval of the respondent are read out and displayed on the display device 58 (step S7). For example, a screen as shown in FIG. 11 is displayed. The numerical values calculated above are embedded in a table that does not contain numerical values in FIG. In addition, the display includes a start button and an end button used in the next phase and a display for questions (however, blank).

  By performing the processing as described above, the phase processing for performing a basic question is performed as preprocessing.

2. Processing in the Phase for Asking the Main Question Next, processing in the phase for asking the main question will be described with reference to FIGS. For example, when a dialogue as shown in FIG. 12 is performed, the optimal calculation is performed at the timing X. That is, when the utterance of only the questioner is started, the next calculation process between the optimum is performed based on the basic data and the data obtained so far in this phase. Further, the optimum interval D is displayed at the start Y of the silent section after the respondent speaks. For example, the display as shown in FIG. 13 is performed. In FIG. 13, a display for a question (here, 5 seconds) is performed on the screen shown in FIG. 11, and the countdown is performed in units of seconds, for example. You may make it count down by the predetermined unit of less than second. However, as described below, if the answerer or the questioner speaks before the display becomes “0”, the display for the question is cleared.

  When competition occurs, the optimum interval is initialized at the stage of timing Z, so that the adjustment is performed again. In this method, although a time period in which no conflict occurs is set, since a conflict has occurred, initialization is performed and optimization processing is performed from the beginning.

  Specifically, the processes shown in FIGS. 14 to 20 are performed. For example, when the questioner clicks the start button for the main question, the input / output unit 554 instructs the voice related data processing unit 51 and the optimum interval calculation unit 553 to start processing in response to the start instruction, and performs voice processing. The related data processing unit 51 records the audio data from the microphones 1 and 3 (step S41). For example, the voice data digitized as in the first phase is stored in the log data storage unit 53 as it is. Further, the voice related data processing unit 51 sequentially generates records of the interview log data table from the voice data, and stores the records in the log data storage unit 53 (step S43). The processing in this step is basically the same as that in step S3.

  The interview log data table is a data table as shown in FIG. 15, for example. The data format is the same as the interview log basic data table.

  Note that steps S41 and S43 are continuously performed until, for example, the end button is clicked by the questioner.

  On the other hand, when the optimum calculation unit 553 receives the process start time from the input / output unit 554, the inter-optimal calculation unit 553 monitors the record of the interview log data table stored in the log data storage unit 53, and the questioner's speech is started. (Step S45). Specifically, it is determined whether a record of “questioner” as a speaker is generated and a start time is set. For example, a state like the ninth record in FIG. 15 is detected.

  When it is determined that the questioner's speech has been started, the optimal calculation unit 553 performs an optimal calculation process (step S47). The calculation process between the optimum will be described with reference to FIGS.

  First, the inter-optimal calculation unit 553 determines whether it is the first calculation process (step S61). For example, it is determined whether the first record of the interview log data table (FIG. 15) has been generated. In this case, the basic data Basic_T is set in the optimum interval D (step S63). In this way, first, the next interval of the questioner is set using the basic data. Then, the process returns to the original process.

  On the other hand, if it is not the first calculation process, the optimal calculation unit 553 extracts a record of the processing target section based on the setting (step S65). In this embodiment, as shown in FIG. 17B, (1) all records registered in the interview log data table, (2) records from the previous calculation timing to the current calculation timing, and (3) the questioner Or any of the setting ranges set by the user. (2) is for calculating the optimal interval reflecting the previous state, (1) is for calculating the optimal interval reflecting all the states in the question of the subject, (3) It reflects the user's intention, and for example, it is possible to use a setting from the calculation timing N times before to the current calculation timing.

Thereafter, the optimal calculation unit 553 causes the parameter calculation unit 552 to perform parameter calculation processing for the extracted section of the interview log data table (step S67). The basic processing content is the parameter calculation processing described with reference to FIGS. 7 and 10, but the processing target is limited to the interview log data table and the extraction section extracted in step S65. The interview log data table of FIG. 15 is converted into an interview log data table as shown in FIG. 17A. That is, “between respondents”, “respondent”, “between questioners”, and “questioner” are updated and registered for a record in which the speaker is silent. In the example of FIG. 17A, M = 10 between the questioners, T = 10 between the respondents, the speaking rate A = 0.33 (= 10/30), the speaking rate B = 0. 37 (= (5 + 5 + 5) / (10 + 5 + 10 + 5 + 10)), total question time T _q = 30, and total answer time T _a = 40 are calculated.

  Then, the inter-optimal calculation unit 553 performs an inter-adjustment process using the data stored in the parameter storage unit 57 (step S69).

  In the present embodiment, the inter-adjustment process is defined under the following consideration.

  That is, the cognitive thinking process from receiving a question to answering is (1) interpreting the contents of the question, (2) thinking about the answer to it, and (3) speaking the answer.

  People who start thinking and speaking start taking T longer than usual after receiving a question and answering, and if they start talking or think while speaking, the speaking speed will be slower than usual. Appears as an action.

  On the other hand, when respondents speak without taking more time than usual, or when speaking faster than usual, there is a possibility that they have answered without thinking too much. In that case, you may not get a good answer. As a cause, it is conceivable that the respondent is pulled and responding because the questioner is asking questions quickly.

  In such a case (hereinafter referred to as pattern A), the time until the questioner asks questions is increased so that the answerer is not pulled by the questioner.

Moreover, as a reason why the respondent answers faster than usual without being pulled by the questioner, the question content may be simple, or there may be a desire to finish quickly. In such cases, the following may appear as actions.
・ Speaking in a shorter time than usual ・ Speaking in a faster way than usual ・ Speaking in a faster way than the questioner ・ Speaking in a shorter time than the questioner

  In the above case (hereinafter referred to as pattern B), the time until the questioner asks a question is shortened, and the respondent can easily answer by matching the respondent's pace.

  On the other hand, the reason why respondents are responding later than usual is that they are responding carefully. For example, there are cases where it takes time to think about answers to difficult questions. In such a case, the person who asks the next question at a later time does not stop the respondent's thoughts, so there is a possibility that a good answer can be drawn.

If you are asking a difficult question, the following behavior may appear:
・ Since the question details are explained in detail, the questioner's speaking time is longer than the respondent's speaking time. ・ Between respondents is longer than usual. Slower than usual

  In the above case (hereinafter referred to as pattern C), a high-quality answer is derived by lengthening the time until the questioner asks a question and providing time for the respondent to think further.

  If this is summarized in the processing flow, it can be realized in two ways: processing based on time and processing based on speech rate.

  First, FIG. 18 shows a processing flow of the inter-adjustment processing when processing is performed based on time. First, the inter-optimal calculation unit 553 determines whether the current time T of the respondents is different from the respondent Basic_T in the basic data (step S71). If T = Basic_T, there is no need for adjustment, so the most recent optimum interval D is set as the optimum interval D (step S85). Then, the process returns to the original process via the terminal B.

  On the other hand, if T and Basic_T are different, it is determined whether T <Basic_T is satisfied (step S73). In the case of T <Basic_T, the optimum interval calculation unit 553 further calculates (the current speech rate A of the questioner> the current speech rate B of the respondent) and (M <the current time of the questioner M <the respondent's It is determined whether it is T) for the present time (step S79). That is, when the questioner's current speaking speed A is faster than the respondent's current speaking speed B, and the current time M of the questioner is shorter than the current time T of the respondent, the pattern described above It corresponds to A. That is, in step S79, it is determined whether it corresponds to the pattern A.

  When the condition of step S79 is satisfied, the optimal interval calculation unit 553 assumes that the answerer is pulled by the questioner and the interval is shortened, and the interval is increased by D = M × Basic_T / T. And stored in the parameter storage unit 57 (step S77). Since Basic_T> T, when the ratio of Basic_T to T is multiplied by M during the current time of the interrogator, the optimal interval D becomes longer than M during the current period. Then, the process returns to the original process.

  On the other hand, when the condition of step S97 is not satisfied, the optimal interval calculation unit 553 calculates (the respondent's current speech speed B> the respondent's speech speed Basic_B in the basic data) and (the questioner's current speech speed). It is determined whether or not A <the current speaking speed B of the respondent and (the current time of the questioner M> the current time T of the respondent) (step S81). That is, when the respondent's speaking rate B is faster than the normal speaking rate Basic_B, the respondent speaks faster than the questioner, and the interval is shorter than between the questioners, the pattern B described above. It corresponds to. That is, it is determined whether it corresponds to the pattern B.

  If the condition of step S81 is satisfied, the optimal interval calculation unit 553 adjusts the parameter so as to shorten the interval by D = M × T / Basic_T, assuming that the respondent wants to end the interview quickly, and stores the parameters. The data is stored in the unit 57 (step S83). Since Basic_T> T, when the ratio of T to Basic_T is multiplied by M during the current time of the interrogator, the optimal interval D is shorter than M during the current period by that ratio. Then, the process returns to the original process.

  On the other hand, if the condition of step S81 is not satisfied, no adjustment is required and the process proceeds to step S85 via the terminal C.

If T> Basic_T in step S73, the optimum interval calculation unit 553 determines (respondent's current speech speed B <respondent's speech speed Basic_B in the basic data) and (sum of question times T _q > It is determined whether or not the total answer time is T _a ) (step S75). That is, when the respondent's current speaking speed B is slower than the normal speaking speed Basic_B and the question time is long, it corresponds to the pattern C described above. That is, it is determined whether it corresponds to the pattern C.

  If the condition of step S75 is satisfied, the optimal interval calculation unit 553 adjusts the interval so as to be longer by D = M × T / Basic_T, and stores it in the parameter storage unit 57 (step S78). Since T> Basic_T, if the ratio of T to Basic_T is multiplied by M during the current time of the interrogator, the optimal interval D becomes longer than M during the current time by that ratio. Then, the process returns to the original process. On the other hand, when the condition of step S75 is not satisfied, the process proceeds to step S85 via the terminal C.

  By carrying out the processing as described above, the optimum time D to be followed by the questioner is calculated.

  Next, FIG. 19 shows a processing flow of the inter-adjustment process in the case of processing based on the speech rate. First, the inter-optimal calculation unit 553 determines whether the current speaking speed B of the respondent is different from the speaking speed Basic_B of the respondent in the basic data (step S91). If B = Basic_B, there is no need to make any adjustments, so the most recent optimum interval D is set as the optimum interval D (step S105). Then, the process returns to the original process via the terminal D.

  On the other hand, if B and Basic_B are different, it is determined whether B> Basic_B (step S93). In the case of B> Basic_B, the optimum interval calculation unit 553 further calculates (the current speech speed A of the questioner> the current speech speed B of the respondent) and (M <the current time of the questioner M <the respondent's It is determined whether it is T) for the present time (step S99). That is, when the questioner's current speaking speed A is faster than the respondent's current speaking speed B, and the current time M of the questioner is shorter than the current time T of the respondent, the pattern described above It corresponds to A. That is, it is determined whether it corresponds to the pattern A.

  When the condition of step S99 is satisfied, the optimal interval calculation unit 553 assumes that the answerer is pulled by the questioner and the interval is shortened, and the interval is increased by D = M × B / Basic_B. And stored in the parameter storage unit 57 (step S97). Since B> Basic_B, if the ratio of B to Basic_B is multiplied by M for the current time of the interrogator, the optimum interval D becomes longer than M for the present time by that ratio. Then, the process returns to the original process.

  On the other hand, when the condition of step S99 is not satisfied, the optimal interval calculation unit 553 (the respondent's current time T <the respondent in the basic data Basic_T) and (the questioner's current speaking speed A < It is determined whether the current speaking speed B of the respondent and (the current time of the questioner M> the current time T of the respondent) (step S101). That is, if the respondent's current T is shorter than normal, the respondent speaks faster than the questioner, and the time is shorter than between the questioners, it corresponds to the pattern B described above. . That is, it is determined whether it corresponds to the pattern B.

  If the condition of step S101 is satisfied, the optimal interval calculation unit 553 adjusts the parameter so as to shorten the interval by D = M × Basic_B / B, assuming that the respondent wants to end the interview quickly, and stores parameters. The data is stored in the unit 57 (step S103). Since B> Basic_B, if the ratio of Basic_B to B is multiplied by M during the current time of the interrogator, the optimal interval D is shorter than M during the current period by that ratio. Then, the process returns to the original process.

  On the other hand, if the condition of step S101 is not satisfied, no adjustment is necessary and the process proceeds to step S105 via the terminal E.

If B <Basic_B in step S93, the optimal time calculation unit 553 calculates (current time T of respondents> Basic_T among respondents in basic data) and (total question time T _q > response time). (Sum T _a )) (step S95). That is, when the respondent's current time T is longer than Basic_T during the normal time and the question time is longer, it corresponds to the pattern C described above. That is, it is determined whether it corresponds to the pattern C.

  If the condition of step S95 is satisfied, adjustment is made so that the interval is longer by D = M × Basic_B / B, and the result is stored in the parameter storage unit 57 (step S98). Since B <Basic_B, if the ratio of Basic_B to B is multiplied by M during the current time of the interrogator, the optimal interval D becomes longer than M during the current period. Then, the process returns to the original process. On the other hand, when the condition of step S95 is not satisfied, the process proceeds to step S105 via the terminal E.

  By carrying out the processing as described above, the optimum time D to be followed by the questioner is calculated.

  When such a process is performed, step S69 in FIG. 16 is completed, and the process returns to the process in FIG.

  Returning to the description of the processing in FIG. 14, when the calculation process between the optimal times in step S47 is completed, the optimal calculation section 553 determines whether the end of the process is instructed by, for example, the questioner (step S57). If the process end is instructed, the process ends. If the process is not finished, the process returns to step S41.

  If it is not the questioner's speech start in step S45, the optimal interval calculation unit 553 determines whether or not the answerer and the questioner are in a competing state in which they speak at the same time (step S49). In the case of a competitive state, the adjustment between them is not successful, and the optimal interval calculation unit 553 sets Basic_T as the optimal interval D (step S51). Then, control goes to a step S57. Further, if not in a competitive state, the record of the interview log data table stored in the log data storage unit 53 is monitored to determine whether it is after the respondent speaks (step S53). Specifically, after the record of the “answerer” is generated and the end time is set, the state where the record of “silence” is generated and the start time is registered is specified.

  When it is determined that the period after the respondent remarks has been started, the optimal interval calculation unit 553 performs an optimal display process (step S55). The optimal display process will be described with reference to FIG.

  The inter-optimal calculation unit 553 instructs the input / output unit 554 to display the optimal interval D, and the input / output unit 554 displays the designated optimal interval D on the display device 58 (step S111). Further, the display numerical value is counted down in units of seconds, for example (step S113). However, the inter-optimal calculation unit 553 monitors whether a record related to the comment of the respondent or the questioner is added to the interview log data table stored in the log data storage unit 53, and detects the record related to the comment. Then, the input / output unit 554 is notified. When the display numerical value is not zero (step S115: No route) and the notification of the speech is received (step S117: Yes route), the input / output unit 554 sets the display numerical value to zero and displays it on the display device 58 ( Step S119). On the other hand, if the displayed numerical value is zero (step S115: Yes route), the process is terminated. On the other hand, if no speech notification has been received (step S117: No route), the process returns to step S113.

  If the above processing is performed, the questioner can confirm the countdown between the optimal times and conduct the question when it becomes zero. If the respondent speaks before it reaches zero, the respondent It is possible to prevent disturbing the tempo.

  It should be noted that not only numerical display but also a case where the time until a question is shortened may be a display prompting to increase the speaking speed such as “you may speak a little faster”. In the case where the time until the question is extended, a display prompting to slow down the speech rate such as “let's do it a little more slowly” may be displayed.

  Returning to the description of the processing in FIG. 14, when it is determined that the time is not after the respondent remarks, the process returns to step S57.

  By performing the processing as described above, the questioner can find the optimum time and can obtain an appropriate answer from the respondent with a good tempo.

  Although the present embodiment has been described above, the embodiment is not limited to this. For example, the functional block diagram shown in FIG. 1 is an example, and may not necessarily match the actual program module configuration. In addition, the processing flow described above can be modified such that the processing order is changed or the processing is executed in parallel if the processing results are the same.

  As for the screen configuration, the example described above is an example, and different screen configurations for displaying the same contents may be adopted.

  Note that the interview support apparatus is a computer apparatus, and as shown in FIG. 21, for a memory 2501, a CPU 2503, a hard disk drive (HDD) 2505, a display control unit 2507 connected to a display device 2509, and a removable disk 2511. The drive device 2513, the input device 2515, and the communication control unit 2517 for connecting to the network are connected by a bus 2519. An operating system (OS) and an application program for executing the processing in this embodiment are stored in the HDD 2505, and are read from the HDD 2505 to the memory 2501 when executed by the CPU 2503. If necessary, the CPU 2503 controls the display control unit 2507, the communication control unit 2517, and the drive device 2513 to perform necessary operations. Further, data in the middle of processing is stored in the memory 2501 and stored in the HDD 2505 if necessary. In an embodiment of the present technology, an application program for performing the above-described processing is stored in a computer-readable removable disk 2511 and distributed, and installed from the drive device 2513 to the HDD 2505. In some cases, the HDD 2505 may be installed via a network such as the Internet and the communication control unit 2517. Such a computer device realizes various functions as described above by organically cooperating the hardware such as the CPU 2503 and the memory 2501 described above with the OS and necessary application programs.

  The present embodiment can be summarized as follows.

  This interview support method is based on the current time of the respondent or the respondent's current speaking speed, which is the time from the end of the speaker's speech to the start of the respondent's speech within the predetermined time during the first interview, An acquisition step of acquiring the current time of the questioner, which is the time from the end of the speech to the end of the questioner's speech, and the first time before the first interview after detecting the end of the questioner's speech during the first interview The standard of the respondent from the standard data storage unit that stores the standard time between the respondent's standards or the standard speed of the respondents, which is the standard time from the end of the commenter's speech to the start of the respondent's speech collected in the interview of 2. Or between the respondent's current speech rate and the respondent's criteria, or comparing the respondent's current speech rate with the respondent's standard speech rate; A first condition that satisfies a first predetermined condition including a condition that the current time of the respondent and the standard of the respondent do not match, or a condition that the current speaking speed of the respondent and the standard speaking speed of the respondent do not match When the predetermined condition of 2 is satisfied, the current time of the questioner is changed according to the relationship between the current time of the respondent and the standard of the respondent, or the relationship between the current speech rate of the respondent and the reference speech speed of the respondent And a step of presenting, to the questioner, the current time of the questioner after the change when the end of the comment of the respondent is detected during the first interview.

  In this way, if the respondent's current time is out of the standard, or if the respondent's speaking speed is out of the standard, the respondent may not be able to say that he / she has successfully answered the interview. Encourage the questioner to adjust between the questioners because there are many. As a result, an appropriate answer can be drawn from the respondent.

  In addition, in the acquisition step described above, both the current time of the respondent and the current speaking speed of the respondent are acquired, and the total question time of the questioner and the respondent within the predetermined time during the first interview. The total answer time may be further acquired. In this case, the reference data storage unit may store both of the respondent's reference and the reference rate of the respondent. And if the change step described above is longer than the respondent's reference time, or if the respondent's current speaking speed is slower than the respondent's reference speaking speed, the total question time of the questioner is answered A first determination step for determining whether the current response speed of the respondent is longer than the total response time of the respondent and whether the current speech speed of the respondent is slower than the reference speech speed of the respondent or the current time of the respondent is longer than the reference time of the respondent If the condition of the step is met, after the change by multiplying the current ratio of the respondent to the current ratio of the respondent or the ratio of the respondent's standard speaking speed to the current speed of the respondent A step of calculating the interval may be included.

  In this way, if an event that seems to be slower than usual in order to think of an answer has occurred, it can be answered at the respondent's pace by lengthening the distance between the questioners as described above. Like that.

  Furthermore, in the acquisition step described above, the current speaking rate of the questioner may be further acquired. At that time, if the change step described above is shorter than the respondent's criteria, or if the respondent's current speech rate is faster than the respondent's standard speech rate, A second determination step for determining whether the current speech rate of the respondent is slower than the current speech rate of the questioner, and a second judgment step that satisfies the conditions of the second judgment step, And calculating a post-change interval by multiplying the ratio of the respondent's current speech rate to the respondent's current reference rate or the respondent's current speech rate to the reference speech rate.

  In this way, if there is an event that the respondent is considered to be involved in the questioner's pace, answer the respondent with a margin by extending the length between the questioners as described above. Can be encouraged to.

  If the change step described above does not satisfy the condition of the second determination step, the respondent's current speech rate is faster than the respondent's standard speech rate, or the respondent's current speech rate is between the respondent's criteria. The third judgment step for judging whether the current time of the respondent is shorter and the current time of the respondent is shorter than the current time of the questioner and whether the current speech speed of the answerer is faster than the current speech speed of the questioner, and the conditions of the third judgment step are satisfied In this case, a step of calculating between the current time of the questioner and the time after the change by multiplying the ratio of the respondent's current speech to the standard of the respondent or the ratio of the respondent's standard speech speed to the current speed of the respondent May be further included.

  In this way, if the respondent wants to respond quickly because the questioner's pace is slow, if there is an event that seems to be fast, by shortening between the questioners as described above , Be able to answer at the respondent's pace.

  It is possible to create a program for causing the hardware to perform the processing described above, and the program can be read by a computer such as a flexible disk, a CD-ROM, a magneto-optical disk, a semiconductor memory, and a hard disk. It is stored in a possible storage medium or storage device. Note that data being processed is temporarily stored in a storage device such as a computer memory.

  Regarding the above embodiment, the following additional notes are disclosed.

(Appendix 1)
Within a predetermined time during the first interview, the current time of the respondent or the current speech speed of the respondent, which is the time from the end of the speech of the questioner to the start of the respondent's speech, and the end of the respondent's speech An acquisition step of acquiring the current time of the questioner, which is a time until the end of the questioner's speech,
From the end of the questioner's speech to the start of the respondent's speech collected in a second interview prior to the first interview after detecting the end of the questioner's speech during the first interview Read the reference rate between the respondents or the reference rate of the respondent from a reference data storage unit that stores the reference time between the reference values of the respondents or the reference rate of the respondents. Or between the respondent's criteria or comparing the respondent's current speech rate with the respondent's reference speech rate;
The first predetermined condition including the condition that the current time of the respondent and the standard of the respondent do not match is satisfied, or the current speaking speed of the respondent and the reference speaking speed of the respondent do not match When the second predetermined condition including the condition is satisfied, the current time of the questioner is determined as the relationship between the current time of the respondent and the criteria of the respondent, or the current speaking speed of the respondent and the respondent. A change step that changes according to the relationship with the reference speech rate of
Presenting to the questioner the current time of the questioner after the change when detecting the end of the comment of the respondent during the first interview;
An interview support method executed by a computer.

(Appendix 2)
In the obtaining step,
While obtaining both the current time of the respondent and the current speaking speed of the respondent, the total question time of the questioner and the total answer time of the respondent within a predetermined time during the first interview Get further,
In the reference data storage unit, both the standard of the respondent and the reference speech rate of the respondent are stored,
The changing step comprises:
When the current period of the respondent is longer than the standard of the respondent or when the current speaking speed of the respondent is slower than the standard speaking speed of the respondent, the total question time of the questioner is the total answer of the respondent A first determination step of determining whether the respondent's current speech rate is longer than the time and is slower than the respondent's reference speech rate or whether the current time of the respondent is longer than the reference time of the respondent;
If the condition of the first determination step is satisfied, the ratio of the respondent to the current ratio of the respondent to the current ratio of the respondent or the standard of the respondent to the current speaking speed of the respondent Calculating the post-change interval multiplied by the speaking rate ratio;
The interview support method according to appendix 1, including

(Appendix 3)
In the obtaining step,
Further acquiring the current speaking rate of the questioner;
The changing step comprises:
When the current time of the respondent is shorter than the reference time of the respondent or when the current speaking speed of the respondent is faster than the reference speech speed of the respondent, the current time of the respondent is higher than the current time of the questioner. A second determination step of determining whether the current speaking speed of the respondent is longer and slower than the current speaking speed of the questioner;
If the condition of the second determination step is satisfied, the current rate of the respondent relative to the current rate of the respondent relative to the current rate of the respondent Calculating the post-change interval multiplied by the speaking rate ratio;
The interview support method according to appendix 2, including

(Appendix 4)
The changing step comprises:
If the condition of the second determination step is not satisfied, the current speaking speed of the respondent is faster than the standard speaking speed of the respondent or the current period of the respondent is shorter than the standard of the respondent, and the A third determination step of determining whether the current time of the respondent is shorter than the current time of the questioner and whether the current speaking speed of the respondent is faster than the current speaking speed of the questioner;
If the condition of the third determination step is satisfied, the ratio of the respondent to the current ratio of the respondent to the current ratio of the respondent or the standard of the respondent to the current speaking speed of the respondent Calculating the post-change interval multiplied by the speaking rate ratio;
The interview support method according to appendix 3, further including:

(Appendix 5)
Within a predetermined time during the first interview, the current time of the respondent or the current speech speed of the respondent, which is the time from the end of the speech of the questioner to the start of the respondent's speech, and the end of the respondent's speech An acquisition step of acquiring the current time of the questioner, which is a time until the end of the questioner's speech,
From the end of the questioner's speech to the start of the respondent's speech collected in a second interview prior to the first interview after detecting the end of the questioner's speech during the first interview Reads the reference rate between the respondents or the reference rate of the respondent from the reference data storage unit storing the reference rate between the respondents or the reference rate of the respondent, and the current rate of the respondent Or between the respondent's criteria or comparing the respondent's current speech rate with the respondent's reference speech rate;
The first predetermined condition including the condition that the current time of the respondent and the standard of the respondent do not match is satisfied, or the current speaking speed of the respondent and the reference speaking speed of the respondent do not match When the second predetermined condition including the condition is satisfied, the current time of the questioner is determined as the relationship between the current time of the respondent and the criteria of the respondent, or the current speaking speed of the respondent and the respondent. A change step that changes according to the relationship with the reference speech rate of
When the end of the respondent's speech is detected during the first interview, presenting the current time of the questioner after the change to the questioner;
Interview support program to make computer run.

(Appendix 6)
In the obtaining step,
While obtaining both the current time of the respondent and the current speaking rate of the respondent, the total question time of the questioner and the total answer time of the respondent within a predetermined time during the first interview Get further,
In the reference data storage unit, both the standard of the respondent and the reference speech rate of the respondent are stored,
The changing step comprises:
When the current period of the respondent is longer than the standard of the respondent or when the current speaking speed of the respondent is slower than the standard speaking speed of the respondent, the total question time of the questioner is the total answer of the respondent A first determination step of determining whether the respondent's current speaking rate is longer than time and is slower than the respondent's reference speaking rate or whether the current time of the respondent is longer than the respondent's reference;
If the condition of the first determination step is satisfied, the ratio of the respondent to the current ratio of the respondent to the current ratio of the respondent or the standard of the respondent to the current speaking speed of the respondent Calculating the post-change interval multiplied by the speaking rate ratio;
The interview support program described in Appendix 5 including.

(Appendix 7)
In the obtaining step,
Further acquiring the current speaking rate of the questioner;
The changing step comprises:
When the current time of the respondent is shorter than the reference time of the respondent or when the current speaking speed of the respondent is faster than the reference speech speed of the respondent, the current time of the respondent is higher than the current time of the questioner. A second determination step of determining whether the current speaking speed of the respondent is longer and slower than the current speaking speed of the questioner;
If the condition of the second determination step is satisfied, the current rate of the respondent relative to the current rate of the respondent relative to the current rate of the respondent Calculating the post-change interval multiplied by the speaking rate ratio;
The interview support program described in Appendix 6 including.

(Appendix 8)
The changing step comprises:
If the condition of the second determination step is not satisfied, the current speaking speed of the respondent is faster than the standard speaking speed of the respondent or the current period of the respondent is shorter than the standard of the respondent, and the A third determination step of determining whether the current time of the respondent is shorter than the current time of the questioner and whether the current speaking speed of the respondent is faster than the current speaking speed of the questioner;
If the condition of the third determination step is satisfied, the ratio of the respondent to the current ratio of the respondent to the current ratio of the respondent or the standard of the respondent to the current speaking speed of the respondent Calculating the post-change interval multiplied by the speaking rate ratio;
The interview support program according to appendix 7, further including:

(Appendix 9)
Within a predetermined time during the first interview, the current time of the respondent or the current speech speed of the respondent, which is the time from the end of the speech of the questioner to the start of the respondent's speech, and the end of the respondent's speech An acquisition means for acquiring the current time of the questioner, which is a time until the end of the questioner's speech,
From the end of the questioner's speech to the start of the respondent's speech collected in a second interview prior to the first interview after detecting the end of the questioner's speech during the first interview A reference data storage unit for storing a reference time between the respondent's standards or the respondent's reference speech rate;
Read the reference rate of the respondent or the reference rate of the respondent from the reference data storage unit, and compare the current rate of the respondent and the reference rate of the respondent, or the current rate of the respondent And a means for comparing the respondent's reference speech rate;
The first predetermined condition including the condition that the current time of the respondent and the standard of the respondent do not match is satisfied, or the current speaking speed of the respondent and the reference speaking speed of the respondent do not match When the second predetermined condition including the condition is satisfied, the current time of the questioner is determined as the relationship between the current time of the respondent and the criteria of the respondent, or the current speaking speed of the respondent and the respondent. Changing means to change according to the relationship with the reference speech rate of
Means for presenting to the questioner the current time of the questioner after the change when detecting the end of the comment of the respondent during the first interview;
Interview support device having

DESCRIPTION OF SYMBOLS 1,3 Microphone 5 Interview assistance apparatus 51 Voice-related data processing part 53 Log data storage part 55 Control part 57 Parameter storage part 58 Display apparatus 551 Basic data processing part 552 Parameter calculation part 553 Optimal calculation part 554 Input / output part

Claims (9)

Within a predetermined time during the first interview, the current time of the respondent or the current speech speed of the respondent, which is the time from the end of the speech of the questioner to the start of the respondent's speech, and the end of the respondent's speech An acquisition step of acquiring the current time of the questioner, which is a time until the end of the questioner's speech,
From the end of the questioner's speech to the start of the respondent's speech collected in a second interview prior to the first interview after detecting the end of the questioner's speech during the first interview Read the reference rate between the respondents or the reference rate of the respondent from a reference data storage unit that stores the reference time between the reference values of the respondents or the reference rate of the respondents. Or between the respondent's criteria or comparing the respondent's current speech rate with the respondent's reference speech rate;
The first predetermined condition including the condition that the current time of the respondent and the standard of the respondent do not match is satisfied, or the current speaking speed of the respondent and the reference speaking speed of the respondent do not match When the second predetermined condition including the condition is satisfied, the current time of the questioner is determined as the relationship between the current time of the respondent and the criteria of the respondent, or the current speaking speed of the respondent and the respondent. A change step that changes according to the relationship with the reference speech rate of
When the end of the respondent's speech is detected during the first interview, presenting the current time of the questioner after the change to the questioner;
An interview support method executed by a computer. In the obtaining step,
While obtaining both the current time of the respondent and the current speaking rate of the respondent, the total question time of the questioner and the total answer time of the respondent within a predetermined time during the first interview Get further,
In the reference data storage unit, both the standard of the respondent and the reference speech rate of the respondent are stored,
The changing step comprises:
When the current period of the respondent is longer than the standard of the respondent or when the current speaking speed of the respondent is slower than the standard speaking speed of the respondent, the total question time of the questioner is the total answer of the respondent A first determination step of determining whether the respondent's current speaking rate is longer than time and is slower than the respondent's reference speaking rate or whether the current time of the respondent is longer than the respondent's reference;
If the condition of the first determination step is satisfied, the ratio of the respondent to the current ratio of the respondent to the current ratio of the respondent or the standard of the respondent to the current speaking speed of the respondent Calculating the post-change interval multiplied by the speaking rate ratio;
The interview support method according to claim 1, comprising: In the obtaining step,
Further acquiring the current speaking rate of the questioner;
The changing step comprises:
When the current time of the respondent is shorter than the reference time of the respondent or when the current speaking speed of the respondent is faster than the reference speech speed of the respondent, the current time of the respondent is higher than the current time of the questioner. A second determination step of determining whether the current speaking speed of the respondent is longer and slower than the current speaking speed of the questioner;
If the condition of the second determination step is satisfied, the current rate of the respondent relative to the current rate of the respondent relative to the current rate of the respondent Calculating the post-change interval multiplied by the speaking rate ratio;
The interview support method according to claim 2, including: The changing step comprises:
If the condition of the second determination step is not satisfied, the current speaking speed of the respondent is faster than the standard speaking speed of the respondent or the current period of the respondent is shorter than the standard of the respondent, and the A third determination step of determining whether the current time of the respondent is shorter than the current time of the questioner and whether the current speaking speed of the respondent is faster than the current speaking speed of the questioner;
If the condition of the third determination step is satisfied, the ratio of the respondent to the current ratio of the respondent to the current ratio of the respondent or the standard of the respondent to the current speaking speed of the respondent Calculating the post-change interval multiplied by the speaking rate ratio;
The interview support method according to claim 3, further comprising: Within a predetermined time during the first interview, the current time of the respondent or the current speech speed of the respondent, which is the time from the end of the speech of the questioner to the start of the respondent's speech, and the end of the respondent's speech An acquisition step of acquiring the current time of the questioner, which is a time until the end of the questioner's speech,
From the end of the questioner's speech to the start of the respondent's speech collected in a second interview prior to the first interview after detecting the end of the questioner's speech during the first interview Read the reference rate between the respondents or the reference rate of the respondent from a reference data storage unit that stores the reference time between the reference values of the respondents or the reference rate of the respondents. Or between the respondent's criteria or comparing the respondent's current speech rate with the respondent's reference speech rate;
The first predetermined condition including the condition that the current time of the respondent and the standard of the respondent do not match is satisfied, or the current speaking speed of the respondent and the reference speaking speed of the respondent do not match When the second predetermined condition including the condition is satisfied, the current time of the questioner is determined as the relationship between the current time of the respondent and the criteria of the respondent, or the current speaking speed of the respondent and the respondent. A change step that changes according to the relationship with the reference speech rate of
Presenting to the questioner the current time of the questioner after the change when detecting the end of the comment of the respondent during the first interview;
Interview support program to make computer run. In the obtaining step,
While obtaining both the current time of the respondent and the current speaking rate of the respondent, the total question time of the questioner and the total answer time of the respondent within a predetermined time during the first interview Get further,
In the reference data storage unit, both the standard of the respondent and the reference speech rate of the respondent are stored,
The changing step comprises:
When the current period of the respondent is longer than the standard of the respondent or when the current speaking speed of the respondent is slower than the standard speaking speed of the respondent, the total question time of the questioner is the total answer of the respondent A first determination step of determining whether the respondent's current speaking rate is longer than time and is slower than the respondent's reference speaking rate or whether the current time of the respondent is longer than the respondent's reference;
If the condition of the first determination step is satisfied, the ratio of the respondent to the current ratio of the respondent to the current ratio of the respondent or the standard of the respondent to the current speaking speed of the respondent Calculating the post-change interval multiplied by the speaking rate ratio;
The interview support program according to claim 5, including: In the obtaining step,
Further acquiring the current speaking rate of the questioner;
The changing step comprises:
When the current time of the respondent is shorter than the reference time of the respondent or when the current speaking speed of the respondent is faster than the reference speech speed of the respondent, the current time of the respondent is higher than the current time of the questioner. A second determination step of determining whether the current speaking speed of the respondent is longer and slower than the current speaking speed of the questioner;
If the condition of the second determination step is satisfied, the current rate of the respondent relative to the current rate of the respondent relative to the current rate of the respondent Calculating the post-change interval multiplied by the speaking rate ratio;
An interview support program according to claim 6 including: The changing step comprises:
If the condition of the second determination step is not satisfied, the current speaking speed of the respondent is faster than the standard speaking speed of the respondent or the current period of the respondent is shorter than the standard of the respondent, and the A third determination step of determining whether the current time of the respondent is shorter than the current time of the questioner and whether the current speaking speed of the respondent is faster than the current speaking speed of the questioner;
If the condition of the third determination step is satisfied, the ratio of the respondent to the current ratio of the respondent to the current ratio of the respondent or the standard of the respondent to the current speaking speed of the respondent Calculating the post-change interval multiplied by the speaking rate ratio;
The interview support program according to claim 7, further comprising: Within a predetermined time during the first interview, the current time of the respondent or the current speech speed of the respondent, which is the time from the end of the speech of the questioner to the start of the respondent's speech, and the end of the respondent's speech An acquisition means for acquiring the current time of the questioner, which is a time until the end of the questioner's speech,
From the end of the questioner's speech to the start of the respondent's speech collected in a second interview prior to the first interview after detecting the end of the questioner's speech during the first interview A reference data storage unit for storing a reference time between the respondent's standards or the respondent's reference speech rate;
Read the reference rate of the respondent or the reference rate of the respondent from the reference data storage unit, and compare the current rate of the respondent and the reference rate of the respondent, or the current rate of the respondent And a means for comparing the respondent's reference speech rate;
The first predetermined condition including the condition that the current time of the respondent and the standard of the respondent do not match is satisfied, or the current speaking speed of the respondent and the reference speaking speed of the respondent do not match When the second predetermined condition including the condition is satisfied, the current time of the questioner is determined as the relationship between the current time of the respondent and the criteria of the respondent, or the current speaking speed of the respondent and the respondent. Changing means to change according to the relationship with the reference speech rate of
Means for presenting to the questioner the current time of the questioner after the change when detecting the end of the comment of the respondent during the first interview;
Interview support device having

Images