Slide Selection System

Abstract

To provide means that selects suitable slides from a plurality of slides during making a presentation.SOLUTION: A prediction model storage unit 11 is configured to store a prediction model 7 indicative of a relation among characteristics of slide groups, attributes of participants and reactions of the participants. A reaction acquisition unit 14 is configured to give characteristics of respective selectable slide groups and attributes of participants to the prediction model 7 to acquire reactions of the participant. A slide selection unit 15 is configured to calculate an evaluation value on the basis of the reactions of the participants, and select a slide group on the basis of the evaluation value. A reaction detection unit 13 is configured to detect reactions of the participants to whom the slide group for use is presented. A reaction prediction unit 16 is configured to give one or more characteristics of the slide group for use and an attribute of the participants to the prediction model 7, and predict one or more reactions of the participants. An attribute estimation unit 17 is configured to, in accordance with a difference existing between reactions of two participants, give the one or more characteristics of the slide group for use and the detected one or more reactions of the participants to the prediction model 7, and estimate the attribute of the participant.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a slide selection system for selecting a suitable slide from among a plurality of slides. [Background technology]

[0002] Presenters at seminars, presentations, academic conferences, lectures, etc. give presentations by showing multiple slides to the audience in a predetermined order. Skilled presenters skip planned slides or change to other slides depending on the audience's reactions during the presentation. By selecting slides in this way based on the audience's reactions, the quality of the presentation can be improved.

[0003] Patent document 1 describes a presentation device that includes an interest level detection means for detecting the level of interest of participants in a presentation, a selection means for selecting at least one slide progression pattern from a plurality of predetermined slide progression patterns according to the detected level of interest, and a display means for displaying slides in a progression order based on the selected slide progression pattern. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Laid-Open No. 2002-149145 Summary of the Invention [Problem to be solved by the invention]

[0005] It is difficult for presenters with low skills to select slides that reflect the reactions of the audience. Therefore, the quality of a presentation depends on the presenter's skills. Also, in seminars and other events, the content that needs to be conveyed to the audience is predetermined, so there are certain restrictions on the slide selection, and slides cannot be selected freely.

[0006] The present invention has been made in view of the above circumstances, and its object is to provide a means for selecting a suitable slide from among a plurality of slides during a presentation. [Means for solving the problem]

[0007] (1) A slide selection system of the present invention includes a prediction model storage unit that stores a prediction model showing the relationship between the characteristics of a group of slides consisting of a plurality of slides presented in sequence, the attributes of the audience, and the audience's reactions; an attribute input unit for inputting the attributes of the audience of a presentation; a reaction acquisition unit that, for each of a plurality of slides selectable in the presentation, provides the characteristics of the group of slides and the attributes of the audience to the prediction model to acquire the audience's reactions; and a calculation unit that calculates an evaluation value for each of the plurality of slides based on the audience's reactions acquired by the reaction acquisition unit, and calculates the calculated evaluation value. The system includes a slide selection unit that selects a group of slides to be used from the plurality of slide groups based on value; a response detection unit that detects the response of the students to whom the group of slides to be used is presented; a response prediction unit that predicts the response of the students by providing the characteristics of the group of slides to be used and the attributes of the students to the prediction model; and an attribute estimation unit that estimates the attributes of the students by providing the characteristics of the group of slides to be used and the reactions of the students detected by the response detection unit to the prediction model in accordance with a difference between the reactions of the students detected by the response detection unit and the reactions of the students predicted by the response prediction unit.

[0008] According to the slide selection system, slides to be used are selected based on the reactions of the students acquired using a predictive model. Furthermore, if the actual reactions of the students differ from the reactions predicted by the system, the attributes of the students assigned to the predictive model are estimated and changed. Therefore, appropriate slides can be selected according to the reactions of the students during the presentation, improving the quality of the presentation.

[0009] (2) Preferably, the prediction model is obtained by machine learning using training data that includes correspondences between the characteristics of a group of slides, the attributes of the audience, and the audience's reactions for a given presentation.

[0010] With this configuration, a prediction model can be obtained that shows the relationship between the characteristics of the slide group, the attributes of the students, and the reactions of the students.

[0011] (3) Preferably, the present invention may further include a termination processing unit that, after the presentation is completed, updates the prediction model through machine learning using the characteristics of the slides used, the attributes of the audience, and the audience's reactions as training data.

[0012] According to this configuration, by updating the prediction model based on the results of a completed presentation, it is possible to select suitable slides based on the reactions of the audience in subsequent presentations, thereby improving the quality of the presentation.

[0013] (4) Preferably, after the presentation ends, the end processing unit may change the method of calculating the evaluation value based on the reactions of the audience detected by the reaction detection unit.

[0014] According to this configuration, by changing the method of calculating the evaluation value based on the results of the completed presentation and changing the criteria for selecting the slide group, it is possible to select the most appropriate slide group in response to the reactions of the audience in subsequent presentations, thereby improving the quality of the presentation.

[0015] (5) Preferably, the plurality of slide groups correspond to parts that make up the presentation, and the reaction acquisition unit, the slide selection unit, the reaction prediction unit, and the attribute estimation unit may operate when the parts are switched.

[0016] According to this configuration, a group of slides to be used can be selected from a plurality of groups of slides for each part that constitutes a presentation.

[0017] (6) Preferably, the reaction of the student may include one or more of a smile, a laugh, crossed arms, furrowed brows, and a number of nods.

[0018] According to this configuration, smiles, laughter, crossed arms, wrinkles between the eyebrows, number of nods, etc. are used as reactions of the audience, and suitable slides can be selected according to the reactions of the audience during the presentation.

[0019] (7) Preferably, the characteristics of the slide group may include one or more of the number of slides, the number of characters, the number of illustrations, the number of figures, the number of tables, and the number of columns in a graph.

[0020] With this configuration, the number of slides, the number of characters, the number of illustrations, the number of figures, the number of tables, the number of columns in a graph, etc. can be used as characteristics of the slide group to select appropriate slides depending on the reactions of the audience during the presentation. [Effects of the Invention]

[0021] According to the present invention, suitable slides can be selected during a presentation depending on the reactions of the audience. [Brief explanation of the drawings]

[0022] [Figure 1] FIG. 1 is a block diagram showing the configuration of a slide selection system 10 according to an embodiment. [Figure 2] FIG. 2 is a block diagram showing the hardware configuration of the slide selection system 10. As shown in FIG. [Figure 3] FIG. 3 shows the parts and slides that make up a presentation. [Figure 4] FIG. 4 is a flowchart showing the operation of the slide selection system 10. DETAILED DESCRIPTION OF THE INVENTION

[0023] Hereinafter, a slide selection system according to an embodiment of the present invention will be described with reference to the drawings. The embodiment described below is merely an example of the present invention, and it goes without saying that the embodiment of the present invention can be appropriately modified without departing from the spirit of the present invention.

[0024] [Overview of Slide Selection System 10] As shown in Figure 1, the slide selection system 10 of the embodiment includes a prediction model memory unit 11, an attribute input unit 12, a reaction detection unit 13, a reaction acquisition unit 14, a slide selection unit 15, a reaction prediction unit 16, an attribute estimation unit 17, and an end processing unit 18.

[0025] The slide selection system 10 is used when a presenter gives a presentation at a seminar, presentation, academic conference, lecture, or the like. Before giving a presentation, the presenter prepares a plurality of slides to be shown to the audience during the presentation. The plurality of slides is divided into a plurality of slide groups corresponding to parts that make up the presentation. The slide selection system 10 selects an appropriate slide group from the plurality of slide groups as the slide group to be used during the presentation.

[0026] The prediction model storage unit 11 stores a prediction model 7 that indicates the relationship between the characteristics of a group of slides, the attributes of the audience, and the audience's reactions. The attribute input unit 12 is a means for inputting the attributes of the audience of a presentation. The reaction acquisition unit 14 provides the characteristics of the slide group and the attributes of the audience for each of multiple slide groups that can be selected in a presentation to the prediction model 7 to acquire the audience's reactions. The slide selection unit 15 calculates an evaluation value S (described in detail below) for each of the multiple slide groups based on the audience's reactions acquired by the reaction acquisition unit 14, and selects a slide group to use from the multiple slide groups based on the calculated evaluation value S.

[0027] The reaction detection unit 13 detects the reactions of the students when the group of slides used is presented. The reaction prediction unit 16 predicts the reactions of the students by providing the characteristics of the group of slides used and the attributes of the students to the prediction model 7. If there is a difference between the reactions of the students detected by the reaction detection unit 13 and the reactions of the students predicted by the reaction prediction unit 16, the attribute estimation unit 17 provides the characteristics of the group of slides used and the reactions of the students detected by the reaction detection unit 13 to the prediction model 7 to estimate the attributes of the students.

[0028] After the presentation ends, the finalization processing unit 18 updates the prediction model 7 by machine learning using the characteristics of the slides used, the attributes of the attendees, and their reactions as training data. The finalization processing unit 18 also changes the calculation method of the evaluation value S based on the reactions of the attendees detected by the reaction detection unit 13.

[0029] [Hardware configuration of slide selection system 10] 2, the slide selection system 10 is configured by a computer 20. Peripheral devices are connected to the computer 20. In the example shown in FIG. 2, a camera 31, a microphone 32, and a projector 33 are connected to the computer 20 as peripheral devices.

[0030] The computer 20 includes a CPU 21, a RAM 22, a storage unit 23, an input unit 24, and a display unit 25. The storage unit 23 is, for example, a hard disk drive or an SSD drive. The input unit 24 is, for example, a keyboard or a mouse. The display unit 25 is, for example, a liquid crystal display. The storage unit 23 stores a prediction model 7, a program 8, a slide set 9, and the like. The program 8 is a program executed by the CPU 21. The slide set 9 is prepared by the presenter before giving a presentation. When the computer 20 operates as the slide selection system 10, the program 8 stored in the storage unit 23 is copied and transferred to the RAM 22. The CPU 21 executes the program stored in the RAM 22, causing the computer 20 to function as the slide selection system 10. The RAM 22 and the storage unit 23 function as the prediction model storage unit 11.

[0031] The projector 33 presents the slides selected by the slide selection system 10 (slides included in the group of slides to be used) to the audience. This allows the presenter to give a presentation. If the projector 33 is not connected to the computer 20, the slide selection system 10 displays the selected slides on the display unit 25. In this case, the presenter operates the projector 33 to display the same slides as those displayed on the display unit 25.

[0032] [Presentation structure and slide features] As shown in FIG. 3, a presentation is made up of multiple parts. The parts that make up a presentation include, for example, an introductory section, a preparatory section, a main section, an explanatory section, an illustrative section, a supplementary section, and a conclusion section. Each part is associated with a slide group consisting of one or more slides. A slide group may consist of, for example, about three to five slides. The slides included in a slide group are presented to the audience in a predetermined order. Note that the number of slides included in each slide group is not limited to three to five and may be any number.

[0033] The presenter prepares one or more slide groups corresponding to each part. The presenter also prepares multiple slide groups corresponding to one or more parts. In the example shown in FIG. 3, the presenter prepares slide group 1 consisting of one slide corresponding to the introductory section. The presenter also prepares slide group 2A consisting of three slides and slide group 2B consisting of four slides corresponding to the first preparation section. The presenter also prepares slide group 3A consisting of three slides, slide group 3B consisting of five slides, and slide group 3C consisting of four slides corresponding to the second preparation section. The slide selection system 10 selects the slide groups to be presented to the audience during the presentation from among the multiple selectable slide groups prepared corresponding to the parts.

[0034] As described above, the storage unit 23 stores multiple slide groups 9 prepared by the presenter. Each slide group 9 has characteristics such as the number of slides, the number of characters, the number of illustrations, the number of figures, the number of tables, the number of columns in a graph, etc. The characteristics of a slide group 9 are characteristics of the appearance of the slide group 9 itself or of the slides included in the slide group 9. Many of the characteristics of a slide group 9 are unrelated to the content of the slide. The storage unit 23 stores the characteristics (not shown) of each slide group 9 along with multiple slide groups 9. The characteristics of a slide group 9 may be assigned by the presenter by displaying the slide group 9 stored in the storage unit 23 on the display unit 25. Alternatively, the characteristics of a slide group 9 may be automatically assigned by the slide selection system 10 to the slide group 9 stored in the storage unit 23.

[0035] [Attendee attributes] Attendees of a presentation have various attributes. Attendees' attributes include, for example, business people, academics, couples, families with children, university students, people with general curiosity, and people seeking solutions to problems. Selecting slides according to the attributes of the attendees can improve the quality of the presentation. Attendees' reactions during a presentation also differ depending on their attributes. For example, couples and families with children will chuckle if they receive a good presentation. On the other hand, academics will nod more frequently if they receive a good presentation.

[0036] The attribute input unit 12 is a means for inputting the attributes of the audience of the presentation. The attributes of the audience input via the attribute input unit 12 are stored in the RAM 22 and are referenced when the response acquisition unit 14 acquires the audience's responses and when the response prediction unit 16 predicts the audience's responses.

[0037] [Reactions from participants] The audience reacts in various ways as the presentation progresses. For example, the audience may laugh, smile, cross their arms, frown, nod, etc. The reaction detection unit 13 detects the audience's reactions during the presentation, i.e., the audience's reactions when the slides used are presented.

[0038] The camera 31 captures images of the attendees attending the presentation and outputs image signals. The microphone 32 captures voices and sounds emitted by the attendees attending the presentation and outputs audio signals.

[0039] The reaction detection unit 13 includes an image processing unit that processes image signals output from the camera 31, and an audio processing unit that processes audio signals output from the microphone 32. The image processing unit detects smiling, crossed arms, furrowed brows, nodding, etc. of the participants based on the image signals acquired by the camera 31. The audio processing unit detects laughter from the participants based on the audio signals acquired by the microphone 32. The reaction detection unit 13 detects laughter, smiling, crossed arms, furrowed brows, and the number of nods as reactions of the participants, and outputs the detection results as numerical data.

[0040] In addition, the image processing unit determines whether the audience is gazing at the presenter or the slides based on the image signal acquired by the camera 31, and acquires the audience's gaze rate based on the result of the determination. The acquired gaze rate is referenced when the calculation method of the evaluation value S is changed in the final processing unit 18.

[0041] The response detection unit 13 may detect the response of the student by a method other than the above. For example, the response detection unit 13 may detect the swaying of the student's body as the student's response based on a signal output from a sensor attached to the chair on which the student sits.

[0042] One or more cameras 31 may be connected to the computer 20. One or more microphones 32 may be connected to the computer 20. The more cameras 31 or microphones 32 are connected, the more accurately the reactions of the students can be detected.

[0043] [Prediction Model 7] The prediction model 7 stored in the prediction model storage unit 11 is a model that indicates the relationship between the characteristics of a group of slides, the attributes of the audience, and the audience's responses. The prediction model 7 is obtained by machine learning using training data that includes correspondences between the characteristics of a group of slides, the attributes of the audience, and the audience's responses for presentations that have already been given. The prediction model 7 may be obtained, for example, by causing a neural network to perform deep learning. Any algorithm, such as backpropagation, may be used for the machine learning. The prediction model 7 may be obtained by machine learning in the slide selection system 10 (computer 20), or may be obtained by machine learning in another computer. In the latter case, the prediction model obtained in the other computer is input to the slide selection system 10 via, for example, a communication unit (not shown) of the computer 20 and stored in the prediction model storage unit 11.

[0044] [Operation of slide selection system 10] The operation of the slide selection system 10 will be described with reference to Figure 4. When a presenter gives a presentation, the CPU 21 performs the process shown in Figure 4. Before the CPU 21 performs the process shown in Figure 4, the storage unit 23 stores a plurality of slide groups 9 (see Figure 2) prepared by the presenter and attributes (not shown) of each slide group. The plurality of slide groups 9 stored in the storage unit 23 include slides that correspond to the same part among the plurality of parts that make up the presentation.

[0045] As shown in FIG. 4, the CPU 21 first receives the input attributes of the students (S11). The CPU 21 executing S11 functions as the attribute input unit 12. The presenter may, for example, recognize the attributes of the students based on a list of students, or may estimate the attributes of the students from the appearance of the students who have actually gathered. The presenter starts the presentation after inputting the attributes of the students into the slide selection system 10. Note that the attributes of the students may also be input by someone other than the presenter. While the presentation is being given, the reaction detection unit 13 detects the reactions of the students attending the presentation.

[0046] Next, the CPU 21 determines whether it is time to select a slide (S12). In S12, the CPU 21 determines that it is time to select a slide if, for example, it is time to switch parts of the presentation. If the CPU 21 determines that it is time to select a slide (S12: Yes), the process proceeds to S13. If the CPU 21 determines that it is not time to select a slide (S12: No), the process proceeds to S19 without executing S13 to S18.

[0047] In S13, the CPU 21 obtains the reactions of the students by providing the prediction model 7 with the characteristics of the slide group and the attributes of the students for each of the multiple slide groups that are selectable at that time. The CPU 21 that executes S13 functions as the reaction acquisition unit 14. The prediction model 7 indicates the relationship between the characteristics of the slide group, the attributes of the students, and the reactions of the students. Therefore, by providing the prediction model 7 with the characteristics of the slide group and the attributes of the students, the reactions of the students can be obtained.

[0048] Next, the CPU 21 calculates an evaluation value for each of the plurality of selectable slide groups based on the reactions of the audience acquired in S13 (S14). In S14, the CPU 21 calculates an evaluation value S for each of the slide groups according to the following formula (1). S=X1·A1+X2·A2+X3·A3+X4·A4+X5·A5 …(1) In equation (1), A1 to A5 are numerical data indicating the reactions of the audience acquired in S13, where A1 is laughter, A2 is smiling, A3 is furrowed brow, A4 is folded arms, and A5 is the number of nods. X1 to X5 are coefficients determined before the CPU 21 performs the processing shown in Fig. 4. Note that the coefficients X1 to X5 are changed in S20 after the presentation ends.

[0049] Next, the CPU 21 selects a group of slides to be used from among the plurality of slide groups based on the evaluation value S calculated in S14 (S15). The CPU 21, which executes S14 and S15, functions as the slide selection unit 15. In S15, the CPU 21 selects the slide group with the highest evaluation value S from among the plurality of slide groups as the group of slides to be used.

[0050] Next, CPU 21 provides the characteristics of the previously selected group of slides and the attributes of the students to prediction model 7 to predict the reactions of the students (S16). CPU 21 executing S16 functions as reaction prediction unit 16. As in S13, by providing the characteristics of the previously selected group of slides and the attributes of the students to prediction model 7, the reactions of the students can be predicted. In S16, the reactions of the students to whom the previously selected group of slides has been presented are predicted.

[0051] Next, CPU 21 determines whether there is a difference between the reactions of the students detected by reaction detection unit 13 and the reactions of the students predicted in S16 (S17). In S17, CPU 21 compares the reactions of the students detected by reaction detection unit 13 with the reactions of the students predicted in S16, and determines that there is a difference if the difference between the two exceeds a predetermined threshold. If CPU 21 determines that there is a difference (S17: Yes), it proceeds to S18. If CPU 21 determines that there is no difference (S17: No), it proceeds to S19 without executing S18.

[0052] In S18, the CPU 21 provides the prediction model 7 with the characteristics of the previously selected slide group and the reactions of the students detected by the reaction detection unit 13 to estimate the attributes of the students (S18). The CPU 21, which executes S18 and S19, functions as the attribute estimation unit 17. As described above, the prediction model 7 indicates the relationship between the characteristics of the slide group, the attributes of the students, and the reactions of the students. Therefore, by providing the prediction model 7 with the characteristics of the previously selected slide group and the attributes of the students, the attributes of the students can be estimated. The attributes of the students estimated in S18 are referenced when acquiring the reactions of the students in S13 and when predicting the reactions of the students in S16.

[0053] When the presentation reaches a certain part, if there is a difference between the reactions of the students detected by the reaction detection unit 13 and the reactions of the students predicted in S16, it is estimated that there is an error in the attributes of the students being used at that time. Therefore, before the presentation proceeds to the next part, in S18, the attributes of the students are estimated using the prediction model 7. By using the estimated attributes of the students in subsequent processing, it is possible to obtain the reactions of the students with high accuracy in S13 and to select an appropriate group of slides to be used in S15.

[0054] 4, CPU 21 performs steps S13 to S18 upon determining that it is time to select a slide. The slide selection timing also corresponds to the timing to switch parts. In this manner, reaction acquisition unit 14, slide selection unit 15, reaction prediction unit 16, and attribute estimation unit 17 operate when switching parts that make up a presentation.

[0055] Next, the CPU 21 determines whether the presentation has ended (S19). In S19, the CPU 21 determines that the presentation has ended, for example, after the presenter issues an instruction to end the presentation. If the CPU 21 determines that the presentation has ended (S19: Yes), the process proceeds to S20. If the CPU 21 determines that the presentation has not ended (S19: No), the process proceeds to S12. In this case, the CPU 21 performs the processes from S12 to S18 for the next part of the presentation.

[0056] In S20, CPU 21 performs termination processing. CPU 21 executing S20 functions as termination processing unit 18. In S20, CPU 21 updates prediction model 7 through machine learning using the characteristics of the slides used in the completed presentation, the attributes of the audience, and the audience's reactions as training data. This allows prediction model 7 to be obtained that more accurately indicates the relationship between the characteristics of the slides, the attributes of the audience, and the audience's reactions. Therefore, it is possible to select suitable slides for subsequent presentations and improve the quality of the presentation.

[0057] Furthermore, in S20, the CPU 21 changes the calculation method of the evaluation value S based on the reactions of the students detected by the reaction detection unit 13. In S20, the CPU 21 uses the following formula (2) obtained by substituting the gaze rate T detected by the reaction detection unit 13 for the evaluation value S of j in formula (1). T=X1·A1+X2·A2+X3·A3+X4·A4+X5·A5 …(2)

[0058] The storage unit 23 stores the values ​​A1 to A5 and the gaze rate T for presentations that have already been given, in association with each other. The CPU 21 stores the values ​​A1 to A5 and the gaze rate T for the completed presentation in the storage unit 23. The CPU 21 obtains coefficients X1 to X5 by performing multiple regression analysis on the values ​​A1 to A5 and the gaze rate T stored in the storage unit 23. The obtained coefficients X1 to X5 are referenced when calculating an evaluation value in S14 for the next presentation. In this way, the slide selection system 10 changes the calculation method for the evaluation value S based on the results of the completed presentation, and changes the selection criteria for the slide group. Therefore, in subsequent presentations, it is possible to select an appropriate slide group in accordance with the audience's reactions, thereby improving the quality of the presentation.

[0059] In S20, CPU 21 may identify a group of slides with low evaluation values ​​S regardless of the attributes of the audience, and display the group of slides on display unit 25 as slides to be deleted. When giving a next presentation, the presenter will not use the slides recommended for deletion by slide selection system 10. This will improve the quality of future presentations.

[0060] [Action and effect] According to the slide selection system 10 of this embodiment, a group of slides to be used is selected based on the reactions of the students acquired using the prediction model 7. Furthermore, if the reactions of the students actually detected differ from the reactions of the students predicted by the slide selection system 10, the attributes of the students provided to the prediction model 7 are estimated and changed. Therefore, appropriate slides can be selected according to the reactions of the students during the presentation, improving the quality of the presentation.

[0061] Furthermore, by updating the prediction model 7 based on the results of a completed presentation, it is possible to select suitable slides based on the audience's reactions in subsequent presentations, thereby improving the quality of the presentation. The same effect can be achieved by changing the calculation method of the evaluation value S based on the results of the completed presentation and changing the criteria for selecting slide groups. Furthermore, the reaction acquisition unit 14, slide selection unit 15, reaction prediction unit 16, and attribute estimation unit 17 operate when switching parts, so that slide groups to be used can be selected from multiple slide groups for each part.

[0062] [Variations] In the slide selection system 10, the reactions of the students are defined as smiling, laughing, crossing their arms, furrowing their brows, and the number of nods. However, in the slide selection system according to the modified example, the reactions of the students may include one or more of smiling, laughing, crossing their arms, furrowing their brows, and the number of nods. The reactions of the students may also include other elements. In the slide selection system according to the modified example, the evaluation value S is calculated based on an equation obtained by modifying equation (1) depending on the elements included in the reactions of the students. For example, if the reactions of the students include only smiling and laughter, the evaluation value S is calculated according to the following equation (3). S = X1 A1 + X2 A2 …(3)

[0063] In the slide selection system 10, the characteristics of a slide group are the number of slides, the number of characters, the number of illustrations, the number of figures, the number of tables, and the number of columns in a graph, but in the slide selection system according to the modified example, the characteristics of a slide group may include one or more of the number of slides, the number of characters, the number of illustrations, the number of figures, the number of tables, and the number of columns in a graph. The characteristics of a slide group may also include other elements. [Explanation of symbols]

[0064] 7. Prediction Model 9. Slides 10. Slide Selection System 11. Prediction model storage section 12. Attribute input section 13. Reaction detection unit 14. Response acquisition section 15. Slide selection section 16. Reaction prediction section 17... Attribute estimation section 18. Finalization section

Claims

1. a prediction model storage unit that stores a prediction model that indicates the relationship between the characteristics of a group of slides consisting of a plurality of slides presented in sequence, the attributes of the students, and the reactions of the students; an attribute input unit for inputting attributes of attendees of the presentation; a reaction acquisition unit that acquires reactions from the audience by providing the prediction model with the characteristics of each of a plurality of slide groups that can be selected in the presentation and the attributes of the audience; a slide selection unit that calculates an evaluation value for each of the plurality of slide groups based on the reactions of the students acquired by the reaction acquisition unit, and selects a group of slides to be used from the plurality of slide groups based on the calculated evaluation value; and a reaction detection unit that detects reactions of the students to whom the slides to be used are presented; a reaction prediction unit that predicts reactions of the students by providing the characteristics of the group of slides used and the attributes of the students to the prediction model; and an attribute estimation unit that, in response to a difference between the reactions of the students detected by the reaction detection unit and the reactions of the students predicted by the reaction prediction unit, provides the characteristics of the group of slides used and the reactions of the students detected by the reaction detection unit to the prediction model to estimate the attributes of the students.

2. The slide selection system of claim 1, wherein the predictive model is obtained by machine learning using training data that includes correspondences between the characteristics of a group of slides, the attributes of the participants, and their reactions for presentations that have already been conducted.

3. The slide selection system of claim 1 or 2 further comprises a termination processing unit that updates the prediction model after the presentation ends by machine learning using the characteristics of the group of slides used, the attributes of the participants, and their reactions as training data.

4. 4. The slide selection system according to claim 3, wherein the end processing unit changes the method of calculating the evaluation value based on the reactions of the audience detected by the reaction detection unit after the presentation ends.

5. the plurality of slides correspond to parts constituting the presentation; 5. The slide selection system according to claim 1, wherein the reaction acquisition unit, the slide selection unit, the reaction prediction unit, and the attribute estimation unit operate when the part is switched.

6. 6. The slide selection system according to claim 1, wherein the reactions of the audience include one or more of a smile, a laugh, crossed arms, furrowed brows, and a number of nods.

7. 7. The slide selection system of claim 1, wherein the characteristics of the slide group include one or more of the number of slides, the number of characters, the number of illustrations, the number of figures, the number of tables, and the number of columns in a graph.

Images