JP7273281B2 - Information output program, information output device and information output method - Google Patents

Description

The present invention relates to an information output program, an information output device and an information output method.

For example, a business operator that provides services to users (hereinafter simply referred to as a business operator) builds and operates an information processing system for providing services to users. Specifically, for example, in order to support the smooth progress of meetings held by users, business operators build information processing systems that automatically create meeting minutes (for example, patent References 1 to 4).

JP 2005-277462 A WO2018/015992 JP 2013-008114 A JP 2017-215931 A

Here, in order to smoothly advance the conference as described above, a facilitator who plays a role of appropriately advancing the conference may participate in each conference. Specifically, the facilitator activates the discussion by, for example, encouraging many participants to make statements, and brings the discussion closer to convergence by making necessary statements at appropriate timings. This allows the user to enhance the effectiveness of the conference to be held.

However, for example, if the number of facilitators is insufficient, the user cannot have facilitators participate in all meetings. Therefore, the user may not be able to enhance the effectiveness of the conference to be held.

Accordingly, in one aspect, an object of the present invention is to provide an information output program, an information output device, and an information output method that make it possible to enhance the effectiveness of a conference.

In one aspect of the embodiment, machine learning of parameters for calculating effect information indicating the effect of holding the conference from situation information indicating the progress of the conference is performed using numerical information calculated from the situation information and the effect and corresponding to a new conference in progress based on the numerical information calculated from the situation information corresponding to the new conference in progress by using the parameters. A computer is caused to execute a process of calculating the effect information corresponding to the new meeting and outputting the effect information corresponding to the new meeting while the new meeting is in progress.

According to one aspect, it is possible to enhance the effectiveness of the conference.

FIG. 1 is a diagram showing the configuration of an information processing system 10. As shown in FIG. FIG. 2 is a diagram for explaining the hardware configuration of the information processing device 1. As shown in FIG. FIG. 3 is a functional block diagram of the information processing apparatus 1. As shown in FIG. FIG. 4 is a flowchart for explaining an outline of information output processing according to the first embodiment. FIG. 5 is a diagram for explaining an outline of information output processing in the first embodiment. FIG. 6 is a diagram for explaining an outline of information output processing in the first embodiment. FIG. 7 is a flowchart for explaining details of information output processing according to the first embodiment. FIG. 8 is a flowchart for explaining details of information output processing according to the first embodiment. FIG. 9 is a flowchart for explaining details of information output processing according to the first embodiment. FIG. 10 is a flowchart for explaining details of information output processing according to the first embodiment. FIG. 11 is a diagram for explaining a specific example of the teacher data 131. As shown in FIG. FIG. 12 is a diagram illustrating a specific example of the text information 134. As shown in FIG. FIG. 13 is a diagram for explaining a specific example of the keyword information 135. As shown in FIG. FIG. 14 is a diagram illustrating a specific example of the numerical information 136 calculated in the processes of S35 and S36. FIG. 15 is a diagram illustrating a specific example of the effect information 138 output in the process of S42. FIG. 16 is a diagram explaining a specific example of the questionnaire information 137. As shown in FIG. FIG. 17 is a diagram for explaining a specific example of the questionnaire information 137. As shown in FIG. FIG. 18 is a diagram illustrating a specific example of the effect information 138a input in the process of S43. FIG. 19 is a diagram illustrating a specific example of the teacher data 131a. FIG. 20 is a diagram illustrating a specific example of the numerical information 136a.

[Configuration of information processing system]
First, the configuration of the information processing system 10 will be described. FIG. 1 is a diagram showing the configuration of an information processing system 10. As shown in FIG.

An information processing system 10 shown in FIG. 1 includes an information processing device 1 (hereinafter also referred to as an information generating device 1), an input device 2, and an output device 3. The information processing system 10 shown in FIG.

The input device 2 is a device including, for example, one or more cameras, one or more audio microphones, one or more weight sensors, and one or more temperature/humidity sensors, and transmits acquired information to the information processing device 1 . Specifically, as shown in FIG. 1, the input device 2 acquires, for example, a video image of each participant 21 and a voice including the content of a statement by each participant 21, and transmits them to the information processing device 1. do. Each camera, each audio microphone, each weight sensor, and each quality temperature sensor included in the input device 2 may all be included in a single device, or may be included in a plurality of devices. It may be something that can be

The output device 3 is, for example, one or more projectors or one or more speakers, and outputs information transmitted from the information processing device 1 . Specifically, as shown in FIG. 1, the output device 3 outputs the information transmitted from the information processing device 1 in a state that each participant 21 can recognize it. The output device 3 may consist of a single device, or may consist of a plurality of devices. Also, the output device 3 may be a device integrated with the input device 2 .

Hereinafter, the input device 2 and the output device 3 are assumed to be devices installed in the conference room 20 as shown in FIG. It is also assumed that the input device 2 is arranged on the desk 11 installed in the conference room 20 . Further, in the conference room 20, participants 21a, 21b, 21c, and 21d (hereinafter collectively referred to as participants 21) are having a conference.

The information processing apparatus 1 is a parameter machine for calculating information indicating the effect of holding the conference (hereinafter also referred to as effect information) from information indicating the progress of the conference (hereinafter also referred to as status information). Learning is performed in advance based on a plurality of teacher data each including numerical information and effect information calculated from the situation information.

After that, the information processing apparatus 1 uses the calculated parameters to obtain the effect corresponding to the new conference based on the numerical information calculated from the situation information corresponding to the new conference being held in the conference room 20. Calculate information. Specifically, the information processing device 1 calculates effect information corresponding to the new meeting based on numerical information calculated from the situation information transmitted from the input device 2, for example.

Furthermore, the information processing device 1 outputs the calculated effect information (effect information corresponding to the new meeting) while the new meeting is in progress. Specifically, the information processing device 1 outputs the calculated effect information from the output device 3, for example.

That is, the information processing apparatus 1 uses, for example, a plurality of teacher data including numerical information corresponding to meetings held in the past and effect information created by the facilitator for the meetings held in the past, so that parameter is generated. Then, when a new conference is being held in the conference room 20, the information processing device 1 uses the generated parameters to obtain an effect corresponding to numerical information input from the input device 2 during the progress of the new conference. Calculate information. After that, the information processing device 1 outputs the calculated effect information from the output device 3 as information necessary for enhancing the feasibility of the ongoing conference.

As a result, the information processing device 1 can make the participants 21 recognize information (effect information) that can improve the feasibility of the conference during the progress of the conference. Therefore, users can increase the effectiveness of meetings in which no facilitator participates.

[Hardware configuration of information processing system]
Next, the hardware configuration of the information processing system 10 will be described. FIG. 2 is a diagram for explaining the hardware configuration of the information processing device 1. As shown in FIG.

The information processing apparatus 1 has a CPU (Computing Processing Unit) 101 which is a processor, a memory 102, an external interface (I/O unit) 103, and a storage medium 104, as shown in FIG. Each unit is connected to each other via a bus 105 .

The storage medium 104 has, for example, a program storage area (not shown) for storing a program 110 for outputting information to be recognized by the conference participants 21 (hereinafter also referred to as information output processing). The storage medium 104 also has a storage unit 130 (hereinafter also referred to as an information storage area 130) that stores information used when performing information output processing, for example. Note that the storage medium 104 may be, for example, an HDD (Hard Disk Drive).

The CPU 101 executes the program 110 loaded from the storage medium 104 to the memory 102 to perform information output processing.

Also, the external interface 103 communicates with the input device 2 and the output device 3, for example.

[Functions of information processing system]
Next, functions of the information processing system 10 will be described. FIG. 3 is a functional block diagram of the information processing apparatus 1. As shown in FIG.

As shown in FIG. 3, the information processing apparatus 1 is configured by organic cooperation between hardware such as the CPU 101 and the memory 102 of the information processing apparatus 1 and the program 110, for example, an information management unit 111 and a learning execution unit. Various functions including a unit 112, an input reception unit 113, an information generation unit 114, an information calculation unit 115, and an output instruction unit 116 are realized.

3, the information processing apparatus 1 includes teacher data 131, parameters 132, situation information 133, text information 134, keyword information 135, numerical value information 136, and questionnaire information 137, for example. , and effect information 138 are stored in the information storage area 130 .

The information management unit 111 stores, for example, pre-generated teacher data 131 in the information storage area 130 . The teacher data 131 includes, for example, numerical information 136 calculated from situation information 133 indicating the progress of meetings held in the past, and effect information 138 indicating the effects of meetings held in the past. good. The status information 133 indicating the progress of the conference held in the past includes, for example, video of the conference captured by a camera during the progress of the conference held in the past, audio of the conference collected by an audio microphone, It may include information corresponding to at least one of the weight of sticky notes and pens measured by a weight sensor, and the temperature and humidity measured by a temperature/humidity sensor. Also, the effect information 138 corresponding to the conference held in the past may be created by the facilitator referring to the numerical information 136 corresponding to the conference held in the past, for example.

The learning execution unit 112 performs machine learning of parameters 132 for calculating effect information 138 indicating the effect of holding the conference from status information 133 indicating the progress of the ongoing conference, using the teacher data 131 as input. by doing.

The input reception unit 113 receives input of status information 133 indicating the progress of the new conference in progress. Specifically, the input receiving unit 113 receives input of the situation information 133 transmitted from the input device 2, for example. The situation information 133 transmitted from the input device 2 includes, for example, video of a meeting captured by a camera during the progress of a new meeting, audio of the meeting collected by an audio microphone, and sticky notes measured by a weight sensor. It may include information corresponding to at least one of the weight of the device, pen, etc., and the temperature and humidity measured by the temperature/humidity sensor.

Information generation unit 114 generates numerical information 136 calculated from situation information 133 received by input reception unit 113 .

The information calculation unit 115 calculates effect information 138 corresponding to the new meeting based on the numerical information 136 generated by the information generation unit 114 by using the parameters 132 calculated by the learning execution unit 112 .

The output instruction unit 116 outputs the effect information 138 calculated by the information calculation unit 115 while the new conference is in progress. Descriptions of the text information 134, the keyword information 135, and the questionnaire information 137 will be given later.

[Outline of the first embodiment]
Next, an outline of the first embodiment will be described. FIG. 4 is a flowchart for explaining an outline of information output processing according to the first embodiment. 5 and 6 are diagrams for explaining an outline of information output processing in the first embodiment.

As shown in FIG. 4, the information processing device 1 waits until the learning execution timing (NO in S1). The learning execution timing may be, for example, the timing at which information to start generating the parameters 132 is input to the information processing apparatus 1 .

Then, when it is time to perform learning (YES in S1), the information processing apparatus 1 calculates the effect information 138 indicating the effect of holding the conference from the situation information 133 indicating the progress of the conference. Learning is performed based on a plurality of teaching data 131 including numerical information 136 calculated from situation information 133 and effect information 138 (S2).

Specifically, as shown in FIG. 5, the information processing apparatus 1 performs machine learning of parameters 132 by inputting teacher data 131 stored in the information storage area 130, for example.

Subsequently, the information processing apparatus 1 uses the parameter 132 machine-learned in the process of S2 to create a new meeting based on the numerical information 136 calculated from the situation information 133 corresponding to the new meeting in progress. Effect information 138 corresponding to is calculated (S3).

After that, the information processing apparatus 1 outputs the effect information 138 calculated in the process of S3 while the new conference is in progress (S4).

Specifically, as shown in FIG. 6, the information processing apparatus 1, for example, uses the numerical information 136 calculated based on the situation information 133 acquired by the input apparatus 2 to determine the effect corresponding to the new conference in progress. The output device 3 is instructed to output the information 138 .

That is, the information processing apparatus 1 uses, for example, a plurality of teacher data 131 including numerical information 136 corresponding to meetings held in the past and effect information created by the facilitator for meetings held in the past, Generate parameters in advance.

Then, when a new conference is being held in the conference room 20, the information processing device 1 uses the generated parameters 132 to correspond to the numerical information 136 input from the input device 2 during the progress of the new conference. Effect information 138 to be calculated. After that, the information processing device 1 outputs the calculated effect information 138 from the output device 3 as information necessary for enhancing the feasibility of the ongoing conference.

As a result, the information processing apparatus 1 can make the participants 21 recognize information (effect information 138) that can improve the feasibility of the conference during the progress of the conference. Therefore, users can increase the effectiveness of meetings in which no facilitator participates.

[Details of the first embodiment]
Next, details of the first embodiment will be described. 7 to 10 are flowcharts for explaining details of information output processing according to the first embodiment. 11 to 20 are diagrams for explaining details of information output processing in the first embodiment.

[Learning execution process]
First, of the information output processing, the processing of performing machine learning of the parameters 132 (hereinafter also referred to as learning execution processing) will be described. FIG. 7 is a flowchart for explaining the learning execution process.

As shown in FIG. 7, the learning execution unit 112 of the information processing device 1 waits until the learning execution timing (NO in S11).

Then, when the learning execution timing has come (YES in S11), the learning execution unit 112 calculates, for example, the effect information 138 indicating the effect of holding the meeting from the status information 133 indicating the progress of the meeting. is performed by inputting a plurality of teacher data stored in the information storage area 130 (S12).

That is, the learning execution unit 112, for example, collects a plurality of teacher data including numerical information 136 calculated from the situation information 133 corresponding to the conference held in the past and effect information 138 corresponding to the conference held in the past. Machine learning of parameters 132 is performed by using 131 as an input. A specific example of the teacher data 131 will be described below.

[Specific example of training data]
FIG. 11 is a diagram for explaining a specific example of the teacher data 131. As shown in FIG.

The teacher data 131 shown in FIG. It has as an item "contents" in which numerical values and sentences corresponding to the included information are stored. Note that, in the teacher data 131 shown in FIG. 11, the information whose “item number” is “1” to “12” corresponds to the numerical information 136 calculated from the situation information 133 corresponding to the conference held in the past. It is information to do. Further, in the teacher data 131 shown in FIG. 11, the information with the "item number" of "13" and "14" is information corresponding to the effect information 138 corresponding to the conference held in the past.

Specifically, in the teacher data 131 shown in FIG. 11, the "contents" of the information whose "item" is "facial expression" (information whose "item number" is "1") includes, for example, smiling participants Information in which "1 (people)", which is the number of participants 21, is stored and the "item" is "age" (information whose "item number" is "2") includes, for example, the number of participants 21 in the "contents". "34", which is the variance of the ages of .

Further, in the teacher data 131 shown in FIG. 11, the "content" of the information whose "item" is "number of men" (information whose "item number" is "3") includes, for example, the number of male participants 21 Information that stores the number of people "6 (persons)" and whose "item" is "number of women" (information whose "item number" is "4") includes, for example, female participants "3 (people)", which is the number of 21, is stored.

In addition, in the "contents" of the information whose "item" is "the number of people seated" (information whose "item number" is "5"), for example, "9 (person)” is stored, and the “content” of information whose “item” is “posture” (information whose “item number” is “6”) includes, for example, the participant 21 leaning forward. "2 (people)", which is the number of people, is stored.

In addition, the "content" of the information whose "item" is "incandescence index" (information whose "item number" is "7") is, for example, a numerical value indicating whether or not the discussion is lively. For example, the number of participants 21 who made a statement is stored in the "contents" of information whose "item" is "conversation index" (information whose "item number" is "8"). A certain "3 (people)" is stored.

In addition, in the "content" of the information whose "item" is "emergence index" (information whose "item number" is "9"), for example, a numerical value indicating whether or not the discussion is converging is " 78” is stored, and the “content” of the information whose “item” is the “number of sticky notes” (information whose “item number” is “10”) includes, for example, sticky notes or sticky notes used in the meeting. "8 (sheets)", which is the number of writing sheets, is stored.

In addition, in the "contents" of information whose "item" is "the number of pens" (information whose "item number" is "11"), for example, the number of pens or writing implements containing pens used in the meeting Information in which a certain "3 (books)" is stored and whose "item" is "discomfort index" (information whose "item number" is "12") includes, for example, a meeting based on temperature and humidity. "70", which is a numerical value indicating the situation in the room 20, is stored.

Furthermore, in the "contents" of information whose "item" is "quality of meeting" (information whose "item number" is "13"), for example, a numerical value indicating the quality of the meeting judged by the facilitator, " 18” is stored, and the “content” of information whose “item” is “recommendation” (information whose “item number” is “14”) includes, for example, the facilitator as a recommendation that is preferably recognized by the participant 21 , which is the content judged by , is stored.

[Input reception processing]
Next, among the information output processes, the process of accepting input of the status information 133 transmitted from the input device 2 (hereinafter also referred to as input acceptance process) will be described. FIG. 8 is a flowchart for explaining input reception processing.

As shown in FIG. 8, the input reception unit 113 of the information processing device 1 waits until it receives the status information 133 from the input device 2 (NO in S21).

When the status information 133 is received from the input device 2 (YES in S21), the information management unit 111 stores the status information 133 received in the process of S21 in the information storage area 130 (S22).

That is, the information management unit 111 sequentially accumulates the status information 133 received from the input device 2 in the information storage area 130 .

[Details of information output processing]
Next, the details of the information output process will be described. 9 and 10 are flowcharts for explaining the details of the information output process.

The information generation unit 114 of the information processing device 1 waits until the information generation timing, as shown in FIG. 9 (NO in S31). The information generation timing may be, for example, the timing for each unit of time (for example, every minute) during the time period during which the conference is being held in the conference room 20 .

Then, when the information generation timing has come (YES in S31), the information generation unit 114, for example, selects the status received within the most recent unit time in the process of S21 among the status information 133 stored in the information storage area 130. The information 133 is specified (S32).

Subsequently, the information generating unit 114 extracts text information about the contents of the video of the situation of the meeting and the contents of the voice recording of the situation of the meeting from the situation information 133 specified in the process of S32. 134 is extracted (S33). Specifically, the information generator 114 extracts the text information 134 about the content shown in the video captured by the camera and the content included in the sound collected by the audio microphone.

That is, the information generation unit 114 extracts information (text information 134) in a state that can be used to generate the numerical information 136 from the content of the video of the meeting situation and the content of the voice recording of the meeting situation. do. A specific example of the text information 134 will be described below.

[Specific example of text information]
FIG. 12 is a diagram illustrating a specific example of the text information 134. As shown in FIG.

The text information 134 shown in FIG. 12 includes "item number" in which the identification information of each text information 134 is stored, and the contents shown in the video of the situation of the meeting or the voice of the situation of the meeting being recorded. It has an item of "content" in which the content to be stored is stored.

Specifically, in the text information 134 shown in FIG. 12, the information whose "item number" is "1" has the "content" as "In recent years, the evolution and development of ICT is remarkable, and ICT is used in daily life, society, industry, It is an essential cornerstone that supports the global economy, and is rapidly expanding its role as a driving force to open up the future.”

In addition, in the text information 134 shown in FIG. 12, the information whose "item number" is "2" has the "contents" of "providing people-friendly, safe, secure and comfortable systems and services, We want to create value.This has been our desire and starting point since our founding." is remembered.

Further, in the text information 134 shown in FIG. 12, "a group of people standing around a table" is stored as the "content" in the information whose "item number" is "3". Description of other information included in FIG. 12 is omitted.

Returning to FIG. 9, the information generator 114 extracts the keyword information 135 from the text information 134 extracted in the process of S33 (S34). A specific example of the keyword information 135 will be described below.

[Specific example of keyword information]
FIG. 13 is a diagram for explaining a specific example of the keyword information 135. As shown in FIG.

The keyword information 135 shown in FIG. 13 has items of “item number” in which identification information of each piece of text information 134 is stored and “keyword” in which a keyword extracted from each piece of text information 134 is stored.

Specifically, in the text information 134 described with reference to FIG. 12, the character strings stored in the "contents" of the information whose "item number" is "1" include "ICT", "development", and "society" as keywords. , 'Industry', 'Life', 'Evolution', 'Drivers', 'Global', 'Today' and 'Future'. Therefore, in the process of S34, the information generation unit 114, for example, sets "ICT", "development", " Extract "Society", "Industry", "Life", "Evolution", "Driving Force", "Global", "Recent" and "Future". For example, as shown in FIG. 13, the information generation unit 114 adds "ICT, development, society, industry, life, evolution, driving force, global , the present, the future.

Further, the character strings stored in the "contents" of the information whose "item number" is "2" in the text information 134 explained with reference to FIG. It includes "system", "service", "provide", "people", "thought", "peace of mind" and "creation". Therefore, in the process of S34, the information generation unit 114, for example, selects "value", "customer", Extract "Safety", "System", "Service", "Provide", "Tachi", "Thought", "Security" and "Creation". Then, for example, as shown in FIG. 13, the information generation unit 114 adds ``value, customer, safety, system, service, provision, ``to the ``keyword'' of the information whose ``item number'' is ``2''. Memorize "thoughts, peace of mind, and creation." Description of other information included in FIG. 13 is omitted.

Returning to FIG. 9, based on the number of keywords included in the keyword information 135 extracted in the process of S34, the information generator 114 calculates numerical information 136 corresponding to the situation information 133 identified in the process of S32 (S35 ).

Specifically, in the process of S35, the information generating unit 114 multiplies the number of keywords included in the keyword information 135 extracted in the process of S34 by a coefficient whose value decreases over time. , numerical information 136 (hereinafter also referred to as first numerical information 136) corresponding to the situation information 133 specified in the processing of S32 is calculated.

That is, a conference in which many keywords appear can generally be judged as a high-quality conference. Moreover, it can be determined that each keyword should appear more frequently in the first half of the conference than in the latter half of the conference because of the need to converge the discussion during the conference. Therefore, the information generation unit 114 calculates the first numerical information 136 indicating whether more keywords appear in the first half of the conference by using a coefficient whose value decreases with the passage of time.

In addition, in the process of S35, the information generation unit 114 determines that the distribution of the keywords included in the keyword information 135 extracted in the process of S34 is the timing before the same conference is being held (the timing when the process of S35 was performed before). ), numerical information 136 corresponding to the situation information 133 specified in the process of S32 (hereinafter also referred to as second numerical information 136) is calculated. Specifically, the information generation unit 114, for example, sets the variance of the keywords included in the keyword information 135 extracted in the process of S34 to the largest value among the variances calculated at the previous timing during the same meeting. The second numerical information 136 is calculated by dividing by the variance.

That is, a conference in which variations in keywords appearing as the conference progresses can generally be judged as a high-quality conference. Therefore, the information generator 114 calculates the second numerical information 136 indicating the degree of convergence of variations in keywords appearing as the conference progresses.

As a result, the information processing device 1 can calculate the numerical information 136 with which the quality of the conference being held in the conference room 20 can be evaluated.

After that, the information generator 114 calculates numerical information 136 other than the numerical information 136 calculated in the process of S35 from the situation information 133 specified in the process of S32 (S36). A specific example of the numerical information 136 calculated in the processes of S35 and S36 will be described below.

[Specific example of numerical information]
FIG. 14 is a diagram illustrating a specific example of the numerical information 136 calculated in the processes of S35 and S36.

The numerical information 136 shown in FIG. 14 is composed of "item number" in which the identification information of each numerical information 136 is stored, "name" in which the name of each numerical information 136 is stored, and "name" in which each numerical information 136 is stored. Numeric value" as an item.

Specifically, the information generating unit 114 identifies the participant 21 whose facial expression is smiling from the image captured by the camera, among the situation information 133 identified in the process of S32. Then, for example, as shown in FIG. 14, the information generation unit 114 converts information whose “name” is “facial expression” (information whose “item number” is “1”) into “numerical value”, and the facial expression is a smiling face. "3 (people)", which is the number of participants 21 identified as As a result, the information processing apparatus 1 can determine that the quality of the conference being held in the conference room 20 is higher, for example, as the number of participants 21 with smiling expressions increases.

In addition, the information generation unit 114 identifies (estimates) the age of each participant 21 from the video captured by the camera in the situation information 133 identified in the process of S32. Then, for example, as shown in FIG. 14, the information generating unit 114 generates each participant 21 Store "43" which is the age variance specified for .

In addition, the information generation unit 114 identifies the male participant 21 from the image captured by the camera in the situation information 133 identified in the process of S32. Then, for example, as shown in FIG. 14, the information generating unit 114 generates the number of male participants as the “numerical value” of the information whose “name” is “number of men” (information whose “item number” is “3”). The number of persons 21, "4 (persons)", is stored.

In addition, the information generation unit 114 identifies the female participant 21 from the video captured by the camera in the situation information 133 identified in the process of S32. Then, for example, as shown in FIG. 14, the information generation unit 114 generates the number of male participants as the “numerical value” of the information whose “name” is “number of women” (information whose “item number” is “4”). The number of persons 21, "4 (persons)", is stored.

In addition, the information generation unit 114 identifies the seated participant 21 from the image captured by the camera in the situation information 133 identified in the process of S32. Then, for example, as shown in FIG. 14 , the information generation unit 114 generates the “numerical value” of the information whose “name” is “the number of people seated” (information whose “item number” is “5”). "3 (people)", which is the number of participants 21 present, is stored. As a result, the information processing device 1 can determine, for example, that the quality of the conference being held in the conference room 20 is higher as the number of participants 21 who are seated is smaller.

In addition, the information generation unit 114 identifies the participant 21 leaning forward from the image captured by the camera in the situation information 133 identified in the process of S32. Then, for example, as shown in FIG. 14 , the information generation unit 114 generates forward leaning posture as the “numerical value” of the information whose “name” is “forward leaning posture” (information whose “item number” is “6”). "3 (people)", which is the number of participants 21 in the posture, is stored. Accordingly, the information processing apparatus 1 can determine that the quality of the conference being held in the conference room 20 is higher as the number of the participants 21 leaning forward increases, for example.

Further, for example, as shown in FIG. 14, the information generation unit 114 calculates in S35 as the "numerical value" of the information whose "name" is "incandescent index" (information whose "item number" is "7"). "10", which is the obtained first numerical information 136, is stored.

In addition, the information generation unit 114 identifies the speaking participant 21 from the sound collected by the audio microphone in the situation information 133 identified in the process of S32. Then, for example, as shown in FIG. 14, the information generation unit 114, as a "numerical value" of information whose "name" is "conversation index" (information whose "item number" is "8"), "6 (people)", which is the number of participants 21 present, is stored. Accordingly, the information processing apparatus 1 can determine that the quality of the conference being held in the conference room 20 is higher, for example, as the number of participants 21 who are speaking increases.

14, the information generation unit 114 calculates in S35 as the "numerical value" of the information whose "name" is "emergence index" (information whose "item number" is "9"). "102", which is the obtained second numerical information 136, is stored.

The information generation unit 114 also identifies the number of tags used during the meeting from the weight measured by the weight sensor in the situation information 133 identified in the process of S32. Specifically, the information generating unit 114 calculates the number of sticky notes used during the meeting by, for example, dividing the weight of the sticky note storage area (not shown) that has decreased since the start of the meeting by the weight of each sticky note. Identify. Then, for example, as shown in FIG. 14, the information generation unit 114 converts the information whose “name” is “the number of sticky notes” (information whose “item number” is “10”) into “numerical value” during the meeting. 20 (sheets), which is the number of sticky notes used in . As a result, the information processing apparatus 1 can determine that the quality of the conference being held in the conference room 20 is higher, for example, as the number of tags used during the conference increases.

The information generation unit 114 also identifies the number of pens used during the meeting from the weight measured by the weight sensor in the situation information 133 identified in the process of S32. Specifically, the information generation unit 114 calculates the number of pens used during the meeting by dividing the weight of the pen storage space (not shown) that has decreased since the start of the meeting by the weight of each pen. Identify. Then, for example, as shown in FIG. 14, the information generating unit 114 converts the information whose “name” is “number of pens” (information whose “item number” is “11”) into “numerical value” during the meeting. Store "2", which is the number of pens used in . Accordingly, the information processing apparatus 1 can determine that the quality of the conference being held in the conference room 20 is higher, for example, as the number of pens used during the conference increases.

Furthermore, the information generating unit 114 identifies the temperature and humidity in the conference room 20 from the temperature and humidity measured by the temperature and humidity sensor in the situation information 133 identified in the process of S32. The information generation unit 114 then calculates a discomfort index indicating how far the specified humidity and temperature are from the appropriate values. After that, the information generating unit 114, for example, as shown in FIG. Remember. Thereby, the information processing device 1 can determine that the quality of the conference being held in the conference room 20 is higher as the discomfort index is lower, for example.

Returning to FIG. 10 , the information calculation unit 115 of the information processing apparatus 1 inputs the numerical information 136 calculated in S35 and S36 by using the parameters 132 machine-learned in the processing of S12 or S44, Effect information 138 corresponding to the conference is calculated (S41).

Then, the output instruction unit 116 of the information processing device 1 outputs the effect information 138 calculated in the process of S41 to the output device 3 (S42). A specific example of the effect information 138 output in the process of S42 will be described below.

[Specific example of effect information output in the process of S42]
FIG. 15 is a diagram illustrating a specific example of the effect information 138 output in the process of S42.

The effect information 138 shown in FIG. 15 includes an "item number" in which the identification information of each effect information 138 is stored, a "name" in which the name of each effect information 138 is stored, and the contents of each effect information 138 are stored. "contents" to be included as items.

Specifically, in the effect information 138 shown in FIG. 15, for the information whose “item number” is “1”, “quality of the meeting” is stored as the “name” and “25” is stored as the “content”. ing.

Further, in the effect information 138 shown in FIG. 15, for the information whose "item number" is "2", "recommendation" is stored as the "name" and "actively discussed" is stored as the "content". It is

As a result, the information processing device 1 can make each participant 21 aware of information such as points to be improved regarding the ongoing conference. Therefore, the information processing device 1 can improve the quality of the ongoing conference.

Returning to FIG. 10, the input receiving unit 113 waits until receiving input of new effect information 138 (hereinafter also referred to as effect information 138a) (NO in S43).

That is, after the conference held in the conference room 20 ends, the input reception unit 113 waits until the facilitator prepares the effect information 138 a for the conference and inputs it to the information processing apparatus 1 .

In this case, the facilitator generates effect information 138a by referring to questionnaire information 137, which is the result of a questionnaire conducted on each participant 21 before and after the conference being held, for example. I do. Specific examples of the questionnaire information 137 and the effect information 138a input in the process of S43 will be described below.

[Specific example of questionnaire information]
16 and 17 are diagrams for explaining specific examples of the questionnaire information 137. FIG. Specifically, FIG. 16 is a diagram illustrating a specific example of questionnaire information 137 (hereinafter also referred to as questionnaire information 137a), which is the result of a questionnaire conducted before the start of the meeting. FIG. 17 is a diagram for explaining a specific example of questionnaire information 137 (hereinafter also referred to as questionnaire information 137b), which is the result of a questionnaire conducted after the meeting.

The questionnaire information 137 shown in FIGS. 16 and 17 includes “item number” in which the identification information of each questionnaire information 137 is stored, “question content” in which the question content of each questionnaire information 137 is stored, and each questionnaire information 137 "Answer" in which an answer corresponding to the content of the question is stored as an item.

Specifically, in the questionnaire information 137a shown in FIG. 16, for the information whose "item number" is "1", "question content" is stored as "What is our company's challenge to innovation?" "XXXX" is stored. In addition, in the questionnaire information 137a shown in FIG. 16, for the information whose "item number" is "2", "What is the future direction of growth in our company?" △△△△” is stored. Further, in the questionnaire information 137a shown in FIG. 16, for the information whose "item number" is "3", "How much do you think it will grow?" 〇〇〇” is stored.

On the other hand, in the questionnaire information 137b shown in FIG. 17, in the information whose "item number" is "1", "question content" is stored as "What is our company's challenge to innovation?" ” is stored. In addition, in the questionnaire information 137b shown in FIG. 17, for the information whose "item number" is "2", "question content" is stored as "What is the future direction of growth in our company?" □□□□" is stored. Further, in the questionnaire information 137b shown in FIG. 17, for the information whose "item number" is "3", "How much do you think it will grow?" ☆☆☆” is stored.

That is, in the questionnaire information 137 shown in FIGS. 16 and 17, information (information whose “item number” is “2”) in which “What is the future direction of growth in our company?” "Answer" and the "Answer" of the information (the information whose "item number" is "3") in which "How much do you think it will grow?" Therefore, the facilitator refers to, for example, the questionnaire information 137 shown in FIGS. , the effect information 138a is created.

As a result, the information processing apparatus 1 can perform machine learning on new teacher data 131 (hereinafter also referred to as teacher data 131a) including the effect information 138a created by the facilitator, as will be described later.

[Specific example of effect information input in the process of S43]
Next, it is a figure explaining the specific example of the effect information 138a input by the process of S43. FIG. 18 is a diagram illustrating a specific example of the effect information 138a input in the process of S43. Effect information 138a shown in FIG. 18 has the same items as the effect information 138 described in FIG.

Specifically, in the effect information 138a shown in FIG. 18, in the information whose "item number" is "1", "quality of the meeting" is stored as the "name" and "28" is stored as the "content". ing.

Further, in the effect information 138a shown in FIG. 18, for the information whose "item number" is "2", "recommendation" is stored as the "name" and "actively discussed" is stored as the "content". It is

Returning to FIG. 10, when the input of the effect information 138a is received (YES in S43), the learning execution unit 112 receives the numerical information 136 calculated in the processes of S35 and S36 and the effect information 138a received in the process of S43. Further machine learning of the parameter 132 is performed by inputting the teacher data 131a including the parameter 132 (S44). A specific example of the teacher data 131a will be described below.

[Specific example of new training data]
FIG. 19 is a diagram illustrating a specific example of the teacher data 131a. Note that the teacher data 131a shown in FIG. 19 has the same items as the teacher data 131 explained in FIG.

Specifically, among the information included in the teacher data 131a shown in FIG. 19, the information having the "item number" of "1" to "12" is stored with the same information as the numerical information 136 described with reference to FIG. ing. Among the information included in the teacher data 131a shown in FIG. 19, the information having the "item numbers" of "13" and "14" stores the same information as the effect information 138a described with reference to FIG. .

That is, the learning execution unit 112 performs machine learning on the teacher data 131a including the effect information 138a created by the facilitator each time the information output process is performed.

Thereby, the information processing device 1 can further improve the accuracy of the parameters 132 for calculating the effect information 138 based on the input numerical information 136 . Therefore, the information processing apparatus 1 can output the effect information 138 more accurately during the progress of the conference.

As described above, the information processing apparatus 1 according to the present embodiment performs machine learning of the parameter 132 for calculating the effect information 138 indicating the effect of holding the conference from the situation information 133 indicating the progress of the conference. This is performed in advance based on a plurality of teacher data 131 each including numerical information 136 calculated from information 133 and effect information 138 .

Then, by using the calculated parameter 132, the information processing apparatus 1 responds to the new meeting based on the numerical information 136 calculated from the situation information 133 corresponding to the new meeting in progress in the meeting room 20. Effect information 138 to be calculated. Thereafter, the information processing device 1 outputs the effect information 138 corresponding to the new conference while the new conference is in progress.

That is, the information processing apparatus 1 uses, for example, a plurality of teacher data 131 including numerical information 136 corresponding to meetings held in the past and effect information 138 created by the facilitator for meetings held in the past. , the parameters 132 are generated in advance. Then, when a new conference is being held in the conference room 20, the information processing device 1 uses the generated parameters 132 to correspond to the numerical information 136 input from the input device 2 during the progress of the new conference. Effect information 138 to be calculated. After that, the information processing device 1 outputs the calculated effect information 138 from the output device 3 as information necessary for enhancing the feasibility of the ongoing conference.

As a result, the information processing apparatus 1 can make the participants 21 recognize information (effect information 138) that can improve the feasibility of the conference during the progress of the conference. Therefore, users can increase the effectiveness of meetings in which no facilitator participates.

In addition, in the process of S35, the information generation unit 114 calculates not only the numerical information 136 of the entire conference being held in the conference room 20, but also the numerical information 136 of each participant 21 (hereinafter, numerical information 136a). can be anything. Then, in the process of S42, the output instruction unit 116 not only outputs the effect information 138 about the entire conference to the output device 3, but also outputs information corresponding to the numerical information 136a about each participant 21 to each participant 21. may be output to a participant terminal (not shown) that can be confirmed. A specific example of the numerical information 136a will be described below.

[Specific example of numerical information about each participant]
FIG. 20 is a diagram illustrating a specific example of the numerical information 136a. FIG. 20 is a diagram illustrating a specific example of numerical information 136a about one participant 21 (hereinafter also referred to as a specific participant 21). The numerical information 136a shown in FIG. 20 has the same items as the numerical information 136 explained in FIG.

Specifically, in the numerical information 136a shown in FIG. 20, for the information whose "item number" is "1", "incandescence index" is stored as the "name" and "2" is stored as the "numerical value". there is

In the numerical information 136a shown in FIG. 20, the information having the "item number" of "2" stores "emergence index" as the "name" and "96" as the "numerical value".

That is, the numerical information 136a shown in FIG. 20 includes only information dependent on a specific participant 21 among the information contained in the numerical information 136 described with reference to FIG.

Then, for example, when the “numerical value” corresponding to the “incandescent index” is below the threshold, the output instruction unit 116 sends information prompting the participant terminal of the specific participant 21 to further speak in the process of S42. may be output. Further, for example, when the “numerical value” corresponding to the “emergence index” is below the threshold, the output instruction unit 116 directs the participant terminal of the specific participant 21 to converge the discussion in the process of S42. It may output information prompting further efforts to increase

This allows users to make meetings in which the facilitator is not participating more effective.

The above embodiments are summarized as follows.

(Appendix 1)
Machine learning of parameters for calculating effect information indicating the effect of holding the conference from situation information indicating the progress of the conference is performed by a plurality of teachers each including numerical information calculated from the situation information and the effect information. act on the data,
calculating the effect information corresponding to the new conference based on the numerical information calculated from the situation information corresponding to the new conference in progress by using the parameter;
outputting the effect information corresponding to the new meeting while the new meeting is in progress;
An information output program characterized by causing a computer to execute processing.

(Appendix 2)
In Appendix 1,
The situation information includes a first keyword shown in a video of the situation of the conference and a second keyword contained in the audio recording of the content of the conference.
An information output program characterized by:

(Appendix 3)
In Appendix 2,
The numerical information is information calculated based on the number of the first keywords and the number of the second keywords,
An information output program characterized by:

(Appendix 4)
In Appendix 3,
The numerical information includes, among the first keywords, the number of keywords displayed in the video corresponding to a first unit time including the time after a predetermined time from the start time of the conference, and the second keywords, A value calculated by multiplying the sum of the number of keywords contained in the speech corresponding to the first unit time and a coefficient that decreases as the predetermined time increases,
An information output program characterized by:

(Appendix 5)
In Appendix 3,
The numerical information includes, among the first keywords, a keyword displayed in the video corresponding to a first unit time including a time after a predetermined time from the start time of the conference, and among the second keywords, the A value calculated by dividing a value indicating variation in keywords included in the speech corresponding to the first unit time by a value indicating the variation corresponding to a second unit time before the first unit time. is
An information output program characterized by:

(Appendix 6)
In Appendix 5,
The second unit time is a unit time that has the largest value indicating the corresponding variation among the unit times before the first unit time.
An information output program characterized by:

(Appendix 7)
In Appendix 1,
the context information includes information indicating the posture of each participant in the conference;
An information output program characterized by:

(Appendix 8)
In Appendix 7,
The numerical information is information indicating the number of participants in the conference who are leaning forward,
An information output program characterized by:

(Appendix 9)
In Appendix 1,
The status information includes information indicating whether each participant in the conference is speaking.
An information output program characterized by:

(Appendix 10)
In Appendix 9,
The numerical information is information indicating the number of people who are speaking among the participants of the conference,
An information output program characterized by:

(Appendix 11)
In Appendix 1,
The status information includes information indicating the total weight of writing instruments that can be used in the meeting, and information indicating the total weight of writing instruments that are not used in the meeting among the writing instruments that can be used in the meeting,
An information output program characterized by:

(Appendix 12)
In Appendix 11,
The numerical information is information indicating the number of writing instruments used in the meeting,
An information output program characterized by:

(Appendix 13)
In Appendix 12,
The number of writing instruments used in the meeting is the total weight of writing instruments that can be used in the meeting minus the total weight of writing instruments that are not used in the meeting, per writing instrument that can be used in the meeting is the number calculated by dividing by the weight of
An information output program characterized by:

(Appendix 14)
In Appendix 1,
The status information includes information indicating the total weight of writing paper that can be used in the meeting, and information indicating the total weight of writing paper that is not used in the meeting among the writing paper that can be used in the meeting. ,
An information output program characterized by:

(Appendix 15)
In Appendix 14,
The numerical information is information indicating the number of writing papers used in the meeting,
An information output program characterized by:

(Appendix 16)
In Appendix 15,
The number of writing papers used in the meeting is the value obtained by subtracting the total weight of writing papers not used in the meeting from the total weight of writing papers that can be used in the meeting. is a number calculated by dividing by the weight per sheet of
An information output program characterized by:

(Appendix 17)
In Appendix 1,
The effect information includes information that prompts the participants of the conference to improve their behavior during the progress of the conference.
An information output program characterized by:

(Appendix 18)
In Appendix 1,
The effect information includes a numerical value indicating the effectiveness of holding the meeting,
An information output program characterized by:

(Appendix 19)
In Appendix 1,
The process of performing the machine learning is performed in response to receiving information indicating that the machine learning is to be performed based on the plurality of teacher data.
An information output program characterized by:

(Appendix 20)
In Supplementary Note 1, further,
storing the received status information in a storage unit in response to receiving the status information corresponding to the new conference in progress;
let the computer do the work,
In the process of calculating the effect information, the effect information corresponding to the new meeting is calculated based on the numerical information calculated from the situation information stored in the storage unit;
An information output program characterized by:

(Appendix 21)
In Appendix 1,
The calculating process is performed for each unit time in the time zone in which the meeting is held,
An information output program characterized by:

(Appendix 22)
In Supplementary Note 1, further,
Machine learning of the parameter based on new teacher data including the numerical information calculated from the situation information corresponding to the new meeting and the effect information corresponding to the new meeting,
An information output program characterized by causing a computer to execute processing.

(Appendix 23)
In Appendix 22,
The process of performing machine learning based on the new training data is performed in response to receiving the effect information corresponding to the new meeting.
An information output program characterized by:

(Appendix 24)
Machine learning of parameters for calculating effect information indicating the effect of holding the conference from situation information indicating the progress of the conference is performed by a plurality of teachers each including numerical information calculated from the situation information and the effect information. a data-based learning execution unit;
an information generation unit that calculates the effect information corresponding to the new conference based on the numerical information calculated from the situation information corresponding to the new conference in progress by using the parameter;
an output instruction unit that outputs the effect information corresponding to the new conference while the new conference is in progress;
An information output device characterized by:

(Appendix 25)
In Appendix 24,
The situation information includes a first keyword shown in a video of the situation of the conference and a second keyword contained in the audio recording of the content of the conference.
An information output device characterized by:

(Appendix 26)
In Appendix 24,
The learning execution unit performs machine learning of the parameter based on new teacher data including the numerical information calculated from the situation information corresponding to the new meeting and the effect information corresponding to the new meeting. I do,
An information output device characterized by:

(Appendix 27)
Machine learning of parameters for calculating effect information indicating the effect of holding the conference from situation information indicating the progress of the conference is performed by a plurality of teachers each including numerical information calculated from the situation information and the effect information. act on the data,
calculating the effect information corresponding to the new conference based on the numerical information calculated from the situation information corresponding to the new conference in progress by using the parameter;
outputting the effect information corresponding to the new meeting while the new meeting is in progress;
An information output method characterized by:

(Appendix 28)
In Appendix 27,
The situation information includes a first keyword shown in a video of the situation of the conference and a second keyword contained in the audio recording of the content of the conference.
An information output method characterized by:

(Appendix 29)
In Supplementary Note 27, further,
Machine learning of the parameter based on new teacher data including the numerical information calculated from the situation information corresponding to the new meeting and the effect information corresponding to the new meeting,
An information output method characterized by:

1: Information processing device 2: Input device 3: Output device 10: Information processing system 11: Desk 20: Conference room 21a: Participant 21b: Participant 21c: Participant 21d: Participant

Claims (22)

Calculating effect information indicating the effect of holding the meeting from the situation information including the first keyword shown in the video of the situation of the meeting and the second keyword included in the audio recording of the content of the meeting. Among the first keywords, the keyword displayed in the video corresponding to the first unit time including the time after a predetermined time from the start time of the meeting, and the second keyword , based on a plurality of teacher data each including numerical information calculated based on variations in keywords included in the speech corresponding to the first unit time and the effect information;
calculating the effect information corresponding to the new conference based on the numerical information calculated from the situation information corresponding to the new conference in progress by using the parameter;
outputting the effect information corresponding to the new meeting while the new meeting is in progress;
An information output program characterized by causing a computer to execute processing. In claim 1,
The numerical information includes, among the first keywords, a keyword displayed in the video corresponding to a first unit time including a time after a predetermined time from the start time of the conference, and among the second keywords, the A value calculated by dividing a value indicating variation in keywords included in the speech corresponding to the first unit time by a value indicating the variation corresponding to a second unit time before the first unit time. is
An information output program characterized by: In claim 2,
The second unit time is a unit time that has the largest value indicating the corresponding variation among the unit times before the first unit time.
An information output program characterized by: In claim 1,
the context information includes information indicating the posture of each participant in the conference;
An information output program characterized by: In claim 4,
The numerical information is information indicating the number of participants in the conference who are leaning forward,
An information output program characterized by: In claim 1,
The status information includes information indicating whether each participant in the conference is speaking.
An information output program characterized by: In claim 6,
The numerical information is information indicating the number of people who are speaking among the participants of the conference,
An information output program characterized by: In claim 1,
The status information includes information indicating the total weight of writing instruments that can be used in the meeting, and information indicating the total weight of writing instruments that are not used in the meeting among the writing instruments that can be used in the meeting,
An information output program characterized by: In claim 8,
The numerical information is information indicating the number of writing instruments used in the meeting,
An information output program characterized by: In claim 9,
The number of writing instruments used in the meeting is the total weight of writing instruments that can be used in the meeting minus the total weight of writing instruments that are not used in the meeting, per writing instrument that can be used in the meeting is the number calculated by dividing by the weight of
An information output program characterized by: In claim 1,
The status information includes information indicating the total weight of writing paper that can be used in the meeting, and information indicating the total weight of writing paper that is not used in the meeting among the writing paper that can be used in the meeting. ,
An information output program characterized by: In claim 11,
The numerical information is information indicating the number of writing papers used in the meeting,
An information output program characterized by: In claim 12,
The number of writing papers used in the meeting is the value obtained by subtracting the total weight of writing papers not used in the meeting from the total weight of writing papers that can be used in the meeting. is a number calculated by dividing by the weight per sheet of
An information output program characterized by: In claim 1,
The effect information includes information that prompts the participants of the conference to improve their behavior during the progress of the conference.
An information output program characterized by: In claim 1,
The effect information includes a numerical value indicating the quality of the meeting,
An information output program characterized by: In claim 1,
The process of performing the machine learning is performed in response to receiving information indicating that the machine learning is to be performed based on the plurality of teacher data.
An information output program characterized by: In claim 1, further:
storing the received status information in a storage unit in response to receiving the status information corresponding to the new conference in progress;
let the computer do the work,
In the process of calculating the effect information, the effect information corresponding to the new meeting is calculated based on the numerical information calculated from the situation information stored in the storage unit;
An information output program characterized by: In claim 1,
The calculating process is performed for each unit time in the time zone in which the meeting is held,
An information output program characterized by: In claim 1, further:
Machine learning of the parameter based on new teacher data including the numerical information calculated from the situation information corresponding to the new meeting and the effect information corresponding to the new meeting,
An information output program characterized by causing a computer to execute processing. In claim 19,
The process of performing machine learning based on the new training data is performed in response to receiving the effect information corresponding to the new meeting.
An information output program characterized by: Calculating effect information indicating the effect of holding the meeting from the situation information including the first keyword shown in the video of the situation of the meeting and the second keyword included in the audio recording of the content of the meeting. Among the first keywords, the keyword displayed in the video corresponding to the first unit time including the time after a predetermined time from the start time of the meeting, and the second keyword , a learning execution unit that performs learning based on a plurality of teacher data each including numerical information calculated based on variations in keywords included in the speech corresponding to the first unit time and the effect information;
an information generation unit that calculates the effect information corresponding to the new conference based on the numerical information calculated from the situation information corresponding to the new conference in progress by using the parameter;
an output instruction unit that outputs the effect information corresponding to the new conference while the new conference is in progress;
An information output device characterized by: Calculating effect information indicating the effect of holding the meeting from the situation information including the first keyword shown in the video of the situation of the meeting and the second keyword included in the audio recording of the content of the meeting. Among the first keywords, the keyword displayed in the video corresponding to the first unit time including the time after a predetermined time from the start time of the meeting, and the second keyword , based on a plurality of teacher data each including numerical information calculated based on variations in keywords included in the speech corresponding to the first unit time and the effect information;
calculating the effect information corresponding to the new conference based on the numerical information calculated from the situation information corresponding to the new conference in progress by using the parameter;
outputting the effect information corresponding to the new meeting while the new meeting is in progress;
An information output method characterized in that a computer executes processing .

Images