Minute book generation device, minute book generation system, minute book generation method, and control program

Abstract

To generate an appropriate minute book.SOLUTION: A minute book creation device (10) comprises: an acquisition unit (11) that acquires conference voice data; a character data generation unit (21) that generates character data from the voice data; and a minute book generation unit (14) that generates a conference minute book using the generated character data. The character data generation unit (21) generates the character data using a voice recognition learning model (22) machine-learned by using teacher data in which a rate for each type of construction terms used for the teacher data is a prescribed rate.SELECTED DRAWING: Figure 2

Description

[Technical field]

[0001] The present invention relates to a minutes taking device for taking minutes of a meeting related to the construction industry. [Background technology]

[0002] A minutes-taking system is known as a system for organizing statements made during a meeting. In general, a minutes-taking system generates character data by performing speech recognition processing on speech data of the meeting, and creates minutes based on the character data.

[0003] Patent document 1 describes a minutes-taking system that applies voice recognition processing to acquired voice data to generate text data, creates minutes of a meeting, and detects statements by a specified speaker or statements containing specified words and phrases from the text data. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] JP 2022-118302 A Summary of the Invention [Problem to be solved by the invention]

[0005] In the minutes-taking system, text data is generated by processing voice data through voice recognition, but there is a problem that the accuracy rate of the voice recognition process is not always as desired. In particular, when taking minutes of a meeting in a specific industry, such as the construction industry, there is a problem that appropriate minutes cannot be taken because words specific to the construction industry cannot be accurately recognized.

[0006] One aspect of the present invention aims to provide a minutes-taking device or the like that generates appropriate minutes in meetings related to the construction industry. [Means for solving the problem]

[0007] In order to solve the above-mentioned problems, one embodiment of the present invention provides a minutes-taking device for meetings related to the construction industry, and includes an acquisition unit that acquires audio data of the meeting, a character data generation unit that generates character data from the audio data, and a minutes-taking unit that uses the generated character data to create minutes of the meeting, wherein the character data generation unit generates the character data using a speech recognition learning model that has been machine-trained using training data in which the proportions of each type of construction term in the training data are set to a predetermined ratio.

[0008] In order to solve the above problem, a minutes-taking system according to one embodiment of the present invention includes the minutes-taking device, an audio capture device including a microphone for capturing audio of the meeting in order to transmit the audio to the minutes-taking device as audio data, and a display device for acquiring and displaying the minutes of the meeting from the minutes-taking device.

[0009] In order to solve the above-mentioned problems, a method for creating minutes of a meeting related to the present invention is a method for creating minutes of a meeting related to the construction industry, comprising: an acquisition step of acquiring audio data of the meeting; a character data generation step of generating character data from the audio data; and a minutes creation step of creating minutes of the meeting using the generated character data, wherein in the character data generation step, the character data is generated using a speech recognition learning model that has been machine-trained using training data in which the proportion of each type of construction term in the training data is a predetermined proportion. Effect of the Invention

[0010] The inventors of the present application have discovered that by using a speech recognition learning model that has been machine-trained using training data in which the proportions of each type of multiple construction terms are set in specific proportions, the error rate in speech recognition processing of speech data related to the construction industry can be reduced.

[0011] According to one aspect of the present invention, minutes are created by generating character data through speech recognition processing using the speech recognition learning model, thereby making it possible to generate minutes with less erroneous conversion from speech data to character data. [Brief description of the drawings]

[0012] [Figure 1] FIG. 1 is an overall schematic diagram showing an overview of a first embodiment of the present invention. [Diagram 2] 1 is a functional block diagram showing a configuration of a main part of a minutes taking device according to a first embodiment of the present invention. [Diagram 3] 1 is a diagram for explaining a learning method of a speech recognition learning model used in the minutes creating device. FIG. [Figure 4] FIG. 2 is a diagram showing an example of training data for a learning model for speech recognition; [Diagram 5] FIG. 2 is a diagram showing an example of minutes created by the minutes creating device; [Figure 6] 11 is a flowchart showing a process flow in the minutes taking device. [Figure 7] FIG. 11 is a functional block diagram showing a configuration of a main part of a minutes taking device according to a second embodiment of the present invention. [Figure 8] FIG. 11 is an overall schematic diagram showing an overview according to a third embodiment of the present invention. [Figure 9] FIG. 11 is a functional block diagram showing a configuration of a main part of a minutes taking device according to a third embodiment of the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0013] [Embodiment 1] An embodiment of the present invention will be described in detail below. The minutes creating device 10 according to this embodiment performs speech recognition processing on acquired speech data using a speech recognition learning model 22, generates character data, and creates minutes.

[0014] The speech recognition learning model 22 is a model that is machine-learned using training data in which the ratio of each type of construction terminology in the training data is a predetermined ratio. The minutes-taking device 10 performs speech recognition processing using this model, thereby enabling accurate creation of minutes of meetings in the construction industry.

[0015] The minutes-taking device 10 may acquire audio data during a meeting and take minutes in real time, or may acquire audio data of a meeting after the meeting has ended and take minutes.

[0016] Fig. 1 shows an example in which the minutes taking device 10 is used. Fig. 1 is an overall schematic diagram showing an example in which the minutes taking device 10 is used. Fig. 1 shows a state in which five people, Participant A to Participant E, are participating in a conference. A voice capturing device 20 including a microphone is placed on a table, and the voice capturing device 20 and the minutes taking device 10 are connected to each other so as to be able to communicate with each other via wire or wirelessly.

[0017] The minutes taking device 10 takes minutes of the meeting from the voice data acquired via the voice acquisition device 20, and displays the taken minutes on the display unit 15.

[0018] The minutes taking device 10 will be described in detail with reference to Fig. 2. Fig. 2 is a functional block diagram showing the main configuration of the minutes taking device 10.

[0019] As shown in FIG. 2, the minutes taking device 10 includes an acquisition unit 11, a voice recognition unit 12, an extraction unit 13, a minutes taking unit 14, and a display unit 15.

[0020] The acquisition unit 11 acquires voice data from an external voice acquisition device 20. Then, the acquired voice data is transmitted to the voice recognition unit 12. Note that the acquisition unit 11 may acquire voice data stored in an external device, instead of acquiring the voice data from the voice acquisition device 20. In other words, the acquisition unit 11 may acquire voice data during a conference in real time or in a state close to real time from the voice acquisition device 20, or may acquire voice data that already exists as the voice of the conference.

[0021] The voice recognition unit 12 includes a character data generation unit 21 and a learning model for voice recognition 22. The voice recognition unit 12 performs voice recognition processing on the voice data transmitted from the acquisition unit 11 using the learning model for voice recognition 22. Then, the character data generation unit 21 generates character data from the result of the voice recognition processing. Details of the learning model for voice recognition 22 will be described later.

[0022] The extraction unit 13 extracts a pre-specified character from the character data generated by the character data generation unit 21. Note that the extraction unit 13 is not an essential component and may be omitted.

[0023] For example, the minutes creating unit 14 extracts phrases indicating when, who, where, what, and how from the character data generated by the character data generating unit 21. This makes it possible to distinguish phrases indicating when, who, where, what, and how from other phrases from the character data, so that the user can easily recognize matters indicating when, who, where, what, and how that are considered important for minutes.

[0024] Furthermore, the extraction unit 13 may extract pre-specified phrases specific to the construction industry from the character data generated by the character data generation unit 21. This makes it possible to distinguish construction-specific phrases from other phrases from the character data, allowing the user to easily recognize construction-specific phrases that are considered important for minutes in the construction industry. Examples of construction-specific terms that are considered important for minutes in the construction industry include the following: "On-site Milestones" "Inspection, intermediate inspection, fire inspection, final inspection" "Pouring concrete, confirmation application, notification submission date, start of construction, completion, handover, framing, electricity supply" "Safety conference, disaster prevention council, weekly regular meetings, general regular meetings"

[0025] The minutes creation unit 14 creates minutes by using the character data generated by the character data generation unit 21. The minutes creation unit 14 may create minutes such that the characters extracted by the extraction unit 13 from the character data generated by the character data generation unit 21 are identifiable from other parts.

[0026] The minutes creating unit 14 may store the created minutes as minutes data in a storage unit (not shown), or may transmit the minutes data to an external device.

[0027] The display unit 15 displays the minutes generated by the minutes taking unit 14. The display unit 15 does not necessarily have to be provided in the minutes taking device 10, and may be configured to display the minutes on an external display device.

[0028] [Learning model for speech recognition] The speech recognition learning model 22 is a model that is machine-learned using training data in which the ratio of each type of construction terminology in the training data is a predetermined ratio. A specific description will be given with reference to Figs. 3 and 4.

[0029] Fig. 3 is a diagram for explaining a learning method of the learning model for voice recognition 22. As shown in Fig. 3, the learning model for voice recognition 22 is learned using as training data 200 a set of voice data extracted from conference data 210, which is voice data of an actual conference, and text data, which is a correct text of the voice data. That is, the training data 200 includes voice data of a conference at an actual construction site. This allows the voice data of a conference at an actual construction site to be used as training data, thereby improving the accuracy of voice recognition processing of voice data of conferences related to the construction industry.

[0030] Fig. 4 is a diagram showing examples included in the teacher data 200. As shown in Fig. 4, the teacher data 200 includes, as multiple types of construction terminology, short sentences related to construction work, frequent expressions in construction-related meetings, and frequent phrases used at construction sites. In this embodiment, the teacher data 200 is used in which the number of data items decreases in the following order: (1) short sentences related to construction work, (2) proper nouns related to construction, (3) frequent expressions in construction-related meetings, and (4) frequent phrases used at construction sites. Examples of these include the following:

[0031] (1) Short sentences related to construction work "This work will be stopped." "Check the location of the house" "Turning single reinforcement into double reinforcement" "Confirm use of the area with the fire department, etc." "Apply for duct materials" (2) Proper nouns related to construction "Kansai Electric is the name of the company." "Amagasaki City is the name of a place." "Aomori City Hall is the name of the organization." (3) Common expressions used in construction-related meetings "We ask for your cooperation in dealing with this issue." "There is a slight discrepancy." "I'll work on it later this week." (4) Common phrases used on construction sites "Perform work in section a" "Construct type b" "Third Floor"

[0032] The inventors of the present application discovered that the error rate of speech recognition processing of speech data related to the construction industry can be reduced by using training data 200 in which the number of data items decreases in the following order: (1) short sentences related to construction work, (2) proper nouns related to construction, (3) frequently used phrases in construction-related meetings, and (4) frequently used phrases at construction sites. Therefore, by using a speech recognition learning model 22 that has been machine-trained using training data 200 in which the number of data items decreases in this order, the error rate of speech recognition can be reduced.

[0033] Fig. 5 shows an example of the minutes 100 displayed on the display unit 15. As shown in Fig. 5, the minutes 100 include an overview section 110 indicating the date, time, location, and participants of the meeting, and a meeting content section 120 which is the content of the meeting. In the meeting content 120, the remarks of each participant are displayed as text.

[0034] As shown in FIG. 5, the entire minutes 100 are not displayed on the display unit 15, but rather the range that can be displayed on the display unit 15 is displayed, and the entire minutes 100 can be checked by scrolling, etc.

[0035] In the displayed text, the words extracted by the extraction unit 13 from the character data, indicating when, who, where, what, and how, may be displayed so as to be distinguishable from other words and words.

[0036] Furthermore, the displayed text may display pre-specified words specific to the construction industry extracted from the character data by the extraction unit 13 in a manner that makes them distinguishable from other words. Here, "XXX" indicated by the component number 121 and "YYY" indicated by the component number 122 are displayed in a manner that makes them distinguishable from other words.

[0037] [Processing flow in the minutes taking device 10] Next, the flow of processing in the minutes taking device 10 will be described with reference to Fig. 6. Fig. 6 is a flowchart showing the flow of processing in the minutes taking device 10.

[0038] 6, in the minutes taking device 10, the acquisition unit 11 acquires voice data (S101, acquisition step). Next, the voice recognition unit 12 performs voice recognition processing on the voice data acquired by the acquisition unit 11 using the voice recognition learning model 22, and generates character data (S102, character data generation step). After that, the minutes taking unit 14 creates the minutes 100 using the generated character data (S103, minutes taking step).

[0039] As described above, the minutes-taking device 10 according to this embodiment takes minutes of a meeting related to the construction industry. It includes an acquisition unit 11 that acquires audio data of the meeting, a character data generation unit 21 that generates character data from the audio data, and a minutes-taking unit 14 that takes minutes of the meeting using the generated character data.

[0040] The character data generation unit 21 generates character data using a learning model 22 for voice recognition that has been machine-trained using training data in which the proportions of each type of construction term in the construction terms used in the training data are set to a predetermined ratio.

[0041] As described above, the inventors of the present application discovered that the misrecognition rate of speech recognition processing of speech data related to the construction industry can be reduced by using the speech recognition learning model 22 that is machine-learned using training data in which the ratio of each type of multiple construction terminology is set to a specific ratio. Therefore, according to the above configuration, speech recognition processing is performed using the speech recognition learning model 22, character data is generated, and minutes are created, so that minutes can be generated with fewer erroneous conversions from speech data to character data.

[0042] [Embodiment 2] Other embodiments of the present invention will be described below. For ease of explanation, the same reference numerals are given to members having the same functions as those described in the above embodiment, and the description thereof will not be repeated.

[0043] In the above-described embodiment, the minutes are created in the minutes-taking device 10 placed in a conference room, etc. In the present embodiment, the conference room, etc. only captures audio and displays the minutes, and the minutes are created in the minutes-taking device 10A connected via a network.

[0044] In this embodiment, a voice capturing device 20, a display device 30, and a transmission / reception device 50 are placed in a conference room or the like, and a minutes taking device 10A is connected to the transmission / reception device 50 so as to be able to communicate with the transmission / reception device 50 via a network. In other words, this embodiment can be said to be a minutes taking system including the voice capturing device 20, the display device 30, the transmission / reception device 50, and the minutes taking device 10A.

[0045] Fig. 7 is a functional block diagram showing the main configuration of the minutes taking device 10A according to this embodiment. As shown in Fig. 7, the configuration of the minutes taking device 10A is the same as that of the minutes taking device 10 described above. The acquisition unit 11 of the minutes taking device 10A acquires voice data transmitted from the transmission / reception device 50, not from the voice acquisition device 20, via the network. Then, the minutes taking unit 14 transmits the created minutes to the transmission / reception device 50 via the network.

[0046] The transmission / reception device 50 transmits the voice data received from the voice acquisition device 20 to the minutes taking device 10A via the network. The transmission / reception device 50 also displays on the display device 30 the minutes received from the minutes taking device 10A via the network.

[0047] The display device 30 is a display that displays information. Note that the display unit may be provided in the transmitting / receiving device 50 instead of the display device 30.

[0048] In this embodiment, the minutes of a meeting are created not at the site such as a conference room but at a location away from the site. This makes it possible to create and display appropriate minutes even if a minutes-creating device is not installed at the site.

[0049] [Embodiment 3] Further, for the sake of convenience, the same reference numerals are given to the members having the same functions as those described in the above embodiment, and the description thereof will not be repeated.

[0050] In the above-described embodiment, the minutes taking device 10 or the minutes taking device 10A performs the voice recognition process and takes the minutes. In the present embodiment, the voice recognition process and the take the minutes are performed by different devices.

[0051] Fig. 8 shows an overall schematic diagram illustrating an overview of this embodiment. As shown in Fig. 8, in this embodiment, a voice acquisition device 20, a display device 30, and a minutes-taking device 10B are arranged in a conference room or the like, and a voice recognition device A is connected to the minutes-taking device 10B so as to be able to communicate with the minutes-taking device 10B via a network. In other words, this embodiment can be said to be a minutes-taking system including the minutes-taking device 10B, the voice acquisition device 20, the display device 30, and the voice recognition device 12A.

[0052] As shown in Fig. 8, in this embodiment, the speech recognition process is performed at a location away from the meeting site, such as a conference room, rather than at the meeting site. This makes it possible to create and display appropriate minutes even when speech recognition process cannot always be performed at the meeting site.

[0053] FIG. 9 is a functional block diagram showing the configuration of the main parts of a minutes taking device 10B and a voice recognition device 12A according to this embodiment.

[0054] 9, the minutes taking device 10B includes an acquisition unit 11, an extraction unit 13, and a minutes taking unit 14. The voice recognition device 12A includes a voice recognition unit 12 and a learning model for voice recognition 22. The minutes taking device 10B and the voice recognition device 12A are connected to each other via a network so as to be able to communicate with each other.

[0055] The acquisition unit 11 acquires the voice data from the voice acquisition device 20 and transmits the acquired voice data to the voice recognition device 12A. The voice recognition device 12A performs voice recognition processing on the voice data transmitted from the minutes taking device 10B using the voice recognition learning model 22, generates character data, and transmits it to the minutes taking device 10B. The minutes taking device 10B takes minutes by using the character data received from the voice recognition device 12A, and displays the minutes on the display device 30. Note that the minutes taking device 10B itself may be provided with a display unit, instead of the display device 30, and the minutes may be displayed on the display unit.

[0056] [Software implementation example] The functions of the minutes-taking devices 10, 10A, 10B (hereinafter referred to as the "devices") can be realized by a program for causing a computer to function as the device, and a program for causing a computer to function as each control block of the device (voice recognition unit 12, extraction unit 13, minutes-taking unit 14).

[0057] In this case, the device includes a computer having at least one control device (e.g., a processor) and at least one storage device (e.g., a memory) as hardware for executing the program. The control device and storage device execute the program to realize each function described in each of the above embodiments.

[0058] The program may be non-transitory and may be recorded in one or more computer-readable recording media. The recording media may or may not be included in the device. In the latter case, the program may be provided to the device via any wired or wireless transmission medium.

[0059] In addition, some or all of the functions of each of the control blocks can be realized by a logic circuit. For example, an integrated circuit in which a logic circuit that functions as each of the control blocks is formed is also included in the scope of the present invention. In addition, the functions of each of the control blocks can be realized by, for example, a quantum computer.

[0060] 〔summary〕 A minutes-taking device according to one embodiment of the present disclosure is a minutes-taking device for meetings related to the construction industry, and includes an acquisition unit that acquires audio data of the meeting, a character data generation unit that generates character data from the audio data, and a minutes-taking unit that creates minutes of the meeting using the generated character data, wherein the character data generation unit generates the character data using a speech recognition learning model that has been machine-learned using training data in which the proportions of each type of construction term in the training data are set to a predetermined ratio.

[0061] The inventors of the present application have discovered that by using a speech recognition learning model that has been machine-trained using training data in which the proportions of each type of multiple construction terms are set in specific proportions, the error rate in speech recognition processing of speech data related to the construction industry can be reduced.

[0062] According to the above configuration, minutes are created by generating character data through speech recognition processing using the speech recognition learning model, thereby making it possible to generate minutes with less erroneous conversion from speech data to character data.

[0063] The minutes-taking device of aspect 2 of the present disclosure is the same as in aspect 1, but includes an extraction unit that extracts words indicating when, who, where, what, and how from the character data generated by the character data generation unit, and the minutes-taking unit makes the words extracted by the extraction unit distinguishable from other words.

[0064] According to the above configuration, it is possible to distinguish from other words and phrases from the character data the words and phrases indicating when, who, where, what and how it was done, so that the user can easily recognize matters that are considered important in the minutes of a meeting, such as when, who, where, what and how it was done.

[0065] In the minutes-taking device of aspect 3 of the present disclosure, in aspect 1 or 2, the extraction unit extracts pre-specified words and phrases specific to the construction industry from the character data generated by the character data generation unit, and the minutes-taking unit makes the words extracted by the extraction unit distinguishable from other words and phrases.

[0066] According to the above configuration, construction-specific words and phrases can be distinguished from other words and phrases from the character data, so that the user can easily recognize construction-specific words and phrases that are considered important for minutes in the construction industry.

[0067] The minutes taking device according to aspect 4 of the present disclosure is any one of aspects 1 to 3, wherein the teacher data used for machine learning of the speech recognition learning model includes speech data of a meeting held at an actual construction site.

[0068] According to the above configuration, since voice data of a meeting at an actual construction site is used as training data, it is possible to improve the accuracy of voice recognition processing of voice data of meetings related to the construction industry.

[0069] In the minutes creation device according to aspect 4 of the present disclosure, in any one of aspects 1 to 4, the teacher data in which the plurality of types of construction terms are used at a specific ratio includes the following data: (1) short sentences related to construction work; (2) proper nouns related to construction; (3) frequently used phrases in construction-related meetings; (4) Common phrases used on construction sites, in that order.

[0070] The inventors of the present application have discovered that the error rate of speech recognition can be reduced by machine learning a speech recognition model using training data in which the number of data items decreases in the following order: short sentences related to construction work, proper nouns related to construction, frequent phrases in construction meetings, and frequent phrases used at construction sites. According to the above configuration, a speech recognition model is used that has been machine-learned using training data in which the number of data items decreases in the following order: short sentences related to construction work, proper nouns related to construction, frequent phrases in construction meetings, and frequent phrases used at construction sites, so that the error rate of speech recognition can be reduced.

[0071] A minutes-taking system according to aspect 6 of the present disclosure includes a minutes-taking device according to any one of aspects 1 to 5, an audio capture device including a microphone for capturing audio of the meeting in order to transmit the audio to the minutes-taking device as audio data, and a display device for capturing and displaying the minutes of the meeting from the minutes-taking device.

[0072] A method for creating minutes of a meeting related to aspect 7 of the present disclosure is a method for creating minutes of a meeting related to the construction industry, and includes an acquisition step of acquiring audio data of the meeting, a character data generation step of generating character data from the audio data, and a minutes creation step of creating minutes of the meeting using the generated character data, wherein in the character data generation step, the character data is generated using a speech recognition learning model that has been machine-trained using training data in which the proportions of each type of construction term in the training data are specified.

[0073] The minutes-taking device according to each aspect of the present disclosure may be realized by a computer. In this case, the control program of the minutes-taking device that causes the computer to operate as each part (software element) of the minutes-taking device to realize the minutes-taking device on a computer, and the computer-readable recording medium on which it is recorded, also fall within the scope of the present invention.

[0074] The present invention is not limited to the above-described embodiments, and various modifications are possible within the scope of the claims. Embodiments obtained by appropriately combining the technical means disclosed in different embodiments are also included in the technical scope of the present invention. [Explanation of symbols]

[0075] 10, 10A, 10B Minute taking device 11 Acquisition Department 12 Voice Recognition Unit 12A Voice recognition device 13 Extraction part 14 Minutes Taking Department 15 Display 20 Audio capture device 21 Character data generation unit 22 Learning Models for Speech Recognition 30 Display device 50 Transmitting and Receiving Device 100 Minutes 200 Teacher Data 210 Conference Data

Claims

1. A device for creating minutes of a meeting related to the construction industry, comprising: an acquisition unit for acquiring audio data of the conference; a character data generating unit for generating character data from the voice data; a minutes creation unit that creates minutes of the meeting by using the generated character data, The character data generation unit of the minutes creation device generates the character data using a voice recognition learning model that has been machine-learned using training data in which the proportion of each type of construction term in the training data is a predetermined ratio.

2. an extraction unit that extracts phrases indicating when, who, where, what, and how from the character data generated by the character data generation unit; The minutes taking device according to claim 1 , wherein the minutes taking section makes the words extracted by the extraction section distinguishable from other words.

3. The extraction unit extracts pre-specified words and phrases specific to the construction industry from the character data generated by the character data generation unit, The minutes-taking device according to claim 2 , wherein the minutes-taking unit makes the term extracted by the extraction unit distinguishable from other terms.

4. The minutes-taking device according to claim 1 , wherein the training data used for machine learning of the speech recognition learning model includes voice data of a meeting held at an actual construction site.

5. The training data with a specific ratio of the above multiple types of construction terms is The number of data is (1) A short text related to construction work, (2) Proper nouns related to construction, (3) Common phrases used in construction-related meetings, (4) Common phrases used on construction sites: The minutes taking device according to claim 1 , wherein the number of participants decreases in the order of number of participants.

6. A minutes creating device according to any one of claims 1 to 5; a voice capture device including a microphone for capturing voice of a conference in order to transmit the voice to the minutes creating device as voice data; a display device that acquires and displays the minutes of the meeting from the minutes taking device; Includes a minutes-taking system.

7. A method for preparing minutes of a meeting related to the construction industry, comprising: acquiring audio data of the conference; a character data generating step of generating character data from the voice data; and a minutes creation step of creating minutes of the meeting by using the generated character data, A method for creating minutes of a meeting, in which the character data generation step generates the character data using a speech recognition learning model that has been machine-trained using training data in which the proportion of each type of construction term in the training data is a predetermined ratio.

8. 2. A control program for causing a computer to function as the minutes-taking device according to claim 1, the control program causing a computer to function as the character data generating section and the minutes-taking section.

Images