KR20180107576A - Learning achievement prediction method and system using visual behavior of learner watching to learning video - Google Patents

Abstract

The present invention relates to a study achievement prediction method using visual behaviors of a learner viewing a learning video and a study achievement prediction apparatus. The study achievement prediction method predicts study achievement for a learning video due to learning immersion from visual behaviors of a leaner viewing a learning video. The study achievement prediction method comprises the steps of: collecting visual behaviors of a learner viewing a learning video; generating behavior data for each visual behavior; generating a graph for generation frequency for each visual behavior of the learner; setting a peak section due to the generation frequency; and predicting a study achievement degree of the leaner.

Description

TECHNICAL FIELD [0001] The present invention relates to a learning achievement prediction method and a learning achievement prediction system using a visual activity of a learner watching a learning video,

The present invention relates to a method for predicting academic achievement using a visual activity of a learner who watches a learning video, and more particularly, to a method for predicting an academic achievement of a learning video according to a learning commitment from a visual activity of a learner watching a learning video And how to predict academic achievement.

Recently, it is not limited to physical space, and it is possible to acquire various learning in an environment where students can learn with their own devices anytime and anywhere. Such online learning is rapidly growing along with the development of information technology, and is utilized not only in school education, but also in the industrial field, taking advantage of cost-effective advantages.

E-learning is often called e-learning, and these e-learning courses are distributed to a large number of learners through the process of analysis, design, and development based on instructional design theory as they design offline courses. In the case of offline face-to-face education, designers and instructors are often the same, giving the opportunity to assess and supplement them through immediate response of learners to the curriculum while practicing on-site.

On the other hand, in the case of online education, the learner 's reaction to the e - learning process once deployed is only a result of the satisfaction and achievement of the process. It can be said that the unidirectional transmission and acceptance by the supplier, in which the cyclical flow of ADDIE model, which is a textbook model of instructional design, is disconnected.

As the interest in big data has been increased in the society in general under the realistic conditions, there is an attempt to apply it to the education field. Numerous digital footprints left by individual learners in the online learning environment are being reexamined by the development of data analysis techniques, and attempts have been made to use them to objectively understand the learners' learning process.

These attempts have become a new discipline of learning analysis. Learning analytics is an attempt to optimize learning by better understanding the context in which learning takes place. It measures, collects, analyzes, and predicts data that occurs in context of learning behaviors and contexts that occur in technology-mediated learning environments, And the 'convergence science literature supporting the decision making of the enemy'.

These learning analytics are not limited to forecasting through data analysis, but are expanding the research flow and interest to a wide range of prescriptions, such as personalized learning, linkage with learning design, and design of teacher support tools. Another feature of the recent flow of learning analysis is the convergence with other disciplines, expanding the scope of the data to be analyzed.

In addition, the field of physiological psychology using body data has been highlighted for the objective measurement of such psychological phenomena. However, these studies have been conducted to measure focused attention with brain activity, heart rate, gaze fixation ratio, pupil response, etc. However, there is no research on visual behavior of learners in learning context State.

The present invention provides an academic achievement prediction method for evaluating the degree of immersion in a learning context of a learner by using a visual action including a movement of a line of sight according to a learner's psychology and a change in the size of a pupil from one eye during a human sense period do.

The present invention provides an academic achievement prediction method for predicting a learner's academic achievement from an evaluated degree of engagement by assessing the degree of engagement in a learning context through a learner's visual behavior.

A method for predicting an academic achievement according to an embodiment includes collecting visual behaviors of a learner watching a learning video; Generating visual activity data for each of the learner's visual behaviors; Generating a graph of an occurrence frequency of each learner's visual behavior using the visual behavior data; Setting a peak interval according to an occurrence frequency in the entire section of the learning moving picture using the generated graph; And analyzing the visual behavior of the learner in the set peak period to predict the degree of academic achievement of the learner.

The step of generating visual action data according to an exemplary embodiment may use a visual action of a learner responded to by at least one learning stimulus of a teaching action or a learning material in a lecture screen provided in accordance with a time flow of the learning moving picture have.

The steps of generating the visual behavior data according to an embodiment include: 1) a gaze fixation frequency and gaze hopping frequency determined by the duration of the left and right eye movements of a learner; 2) a left and right pupil based on pupil diameters in an average learning period of a learner 3) the coordinate value of the lecture screen of the learning video at the time the gaze is located, and 4) the distance between the eyes of the learner considering the change of the perspective and the screen on which the learning video is displayed. Can be considered.

The sizes of the left and right pupils according to the embodiment can be determined by the presentation style of the learning data provided through the learning animation and the cognitive load or emotion change of the learner caused by the presentation method.

The generating of the graph according to an embodiment generates a graph corresponding to each of the average values of the duration of the gaze fixation frequency, the duration of the gaze hopping frequency, and the size of the left and right pupils based on the lecture flow of the learning moving image .

The step of setting a peak interval according to an exemplary embodiment may set specific time zones within the entire interval based on the graph and may set a point indicating a high occurrence frequency of each of the specific time zones to each peak interval have.

According to one embodiment of the present invention, the method for predicting the academic achievement uses visual behaviors including eye movements according to the learner's psychology and pupil size change from the eye, which is one of the human sense periods, The degree of immersion can be evaluated.

According to one embodiment of the present invention, the academic achievement prediction method can predict the academic achievement level of the learner from the evaluated degree of immersion by evaluating the degree of immersion in the learning context through the visual behavior of the learner.

FIG. 1 is a diagram illustrating an overall configuration for predicting a learner's academic achievement according to an exemplary embodiment of the present invention.
FIG. 2 is a diagram for explaining a detailed configuration of an academic achievement prediction apparatus according to an embodiment.
FIG. 3 is a diagram for explaining learning data constituting a learning moving picture according to an embodiment and visual behavior of a learner according to a behavior of a teacher.
4 is a diagram for explaining a method of generating a graph of visual behavior according to an embodiment.
FIG. 5 is a view for explaining a method of setting a peak interval according to the occurrence frequency in the entire section according to an embodiment.
FIG. 6 is a diagram for explaining a process of predicting an academic achievement according to an embodiment.
7 is a flowchart for explaining a method of predicting an academic achievement according to an embodiment.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

FIG. 1 is a diagram illustrating an overall configuration for predicting a learner's academic achievement according to an exemplary embodiment of the present invention.

1, the learning achievement predicting apparatus 101 can predict the learning achievement according to the visual behavior 107 of the learner 106 in cooperation with the user terminal 102 for displaying the learning moving picture 103 . Here, the learning moving image 103 may be an image including the learning data 105 and the instructor 104 for the learner to provide information.

More specifically, the learning achievement predicting device 101 can collect the visual behavior 107 of the learner 106 who watches the learning video 103. To this end, the academic achievement prediction apparatus 101 may collect the visual behavior 107 of the learner 106 as a body organ associated with visual perception of the external stimulus by the learning animation 103. At this time, visual behavior (107) can be 1) active, selectively involving conscious actions to accept external stimuli, and 2) passive and unconscious actions by autonomic nervous system. The learning achievement prediction apparatus 101 can collect visual behaviors including eye movement, pupillary response, and blink of the learner 106 according to conscious behavior and unconscious behavior .

Here, the eye movement stably maintains the image of the object on the fovea in order to capture the external environment in detail, and the motion of the eye movement is largely determined by 1) keeping the object well in the center of the retina Gaze stabilization (gaze stabilization), and gaze shifting (gaze shift) to move the eye toward the object of interest to form the center of the object.

The pupil is a circle-shaped hollow space at the center of the iris, which has the function of controlling the amount of light coming into the eye. The pupil reaction is caused by the contraction of the two muscles of the iris It can mean that the size changes. Here, when pupil contraction is contracted by the stimulation of parasympathetic nerve, the size of the pupil becomes smaller and the pupil size becomes larger when the pupil dilator shrinks by stimulation of the sympathetic nerve. This is because the expansion and contraction of the pupil are independent of each other, and thus the psychophysiological cause of the reaction can be relatively easily found. Moreover, the size of the pupil can be used as a coordinate to grasp the degree of concentration of the learners as the cognitive load level is expressed as positive / negative / neutral response to the stimulus.

The learning achievement predicting apparatus 101 traces the gaze according to the human visual processing characteristic in order to confirm how much the learner 106 concentrates on the lecture contents of the learning video 103, The visual behavior 107 can be collected. For example, the academic achievement prediction apparatus 101 may capture the visual behavior 107 including the position of the eyes, the moving distance, etc. by capturing the infrared rays projected on the eyes of the learner 106 reflected from the lens . Pupilometry can be used to estimate the intensity and mental states of mental activity and to measure changes in pupil diameter and pupil size can do.

That is, in order to measure visual behavior 107 through eye tracking, the present invention can be used to measure (1) type of eye movement (gaze fixation, gaze jump, mixing) and (2) The visual behavior 107 can be measured. Here, the temporal scale means a duration in a specific area, and can indicate 'when' and 'how long' the cognitive processing is performed. And, the spatial scale measures the position, distance, direction and order of gaze fixation and gaze jump, and can show 'where' and 'how'. The measure of frequency can indicate how high it is to show a high correlation with the temporal scale.

The learning achievement prediction apparatus 101 generates visual behavior data for each visual behavior 107 of the learner 106 and then generates a graph of the occurrence frequency of the visual activity of the learner using the generated visual behavior data have. Here, the graph may be a graph that is linearly expressed according to the temporal flow of the learning moving picture. The detailed configuration will be described with reference to FIG.

The academic achievement prediction apparatus 101 can set a peak interval according to the occurrence frequency in the entire section of the learning moving picture using the graph. The detailed configuration will be described with reference to FIG. The learning achievement predicting device 101 can predict the degree of achievement of the learner 106 by analyzing the visual behavior 107 of the learner 106 at the set peak period.

More specifically, the learning achievement predicting device 101 can measure the degree of the learning commitment of the learner 106 through the visual behavior 107. In order to estimate this, it is necessary to 1) visual distortion, 2) concentration, 3) pleasure, 4) control feeling, 5) Curiosity and the like can be used. Specifically, learning commitment occurs when challenges and techniques balance between the borders of boredom and anxiety, which can mean when an action-awareness merge occurs between learning behavior and learning perception.

In addition, learning commitment can be divided into 1) participation, 2) cognitive engagement, 3) cognitive absorption, and 4) immersion according to the immersion method. And the intrinsic motivation of the learner can be enhanced by the satisfaction of the immersion experience itself as it is utilized as the psychological mechanism that triggers the concentration and participation of the learner. In other words, learning commitment can be a positive / psychological factor that improves academic achievement and affects learning outcomes by increasing the desire for learning throughout the learning environment, promoting voluntary and participatory activities.

Thereafter, the present invention can predict the academic achievement of the learner 106 from the degree of learning commitment of the learner 106 through the visual behavior 107. In detail, the academic achievement is the result obtained after the learner who watches the learning video 103 performs a series of learning process, which may mean the goal reaching information of the learner. At this time, in the present invention, in order to measure the academic achievement, the pre-test and the post-test made by the same kind of learner were conducted, and the improvement information of the post-learning knowledge can be calculated through the score difference between the pre- .

Accordingly, the present invention can predict the degree of the learner's academic achievement of the information included in the learning moving image from the visual behavior 107 and the immersion level of the learner 106 viewing the learning animation 103. [

FIG. 2 is a diagram for explaining a detailed configuration of an academic achievement prediction apparatus according to an embodiment.

2, the academic achievement prediction apparatus 201 includes a visual behavior collecting unit 202, a visual behavior data generating unit 203, a graph generating unit 204, a peak interval setting unit 205, (206). ≪ / RTI >

The visual behavior collecting unit 202 can collect visual behaviors of the learner who watches the learning video. The visual action collecting unit 202 can collect visual behaviors of the learner responding to the learning stimulus provided to the learner through the learning animation. More specifically, the learning video may include 1) a slide area for presenting the learning contents, 2) a teaching area for describing the learning contents, and 3) a writing according to the description of the learning contents.

The visual action collecting unit 202 can collect not only the visual behavior of the learner but also the input information through the writing instrument while watching the learning video. In other words, the visual action collecting unit 202 can use an eye tracker, a webcam and a tablet pen in a hardware configuration, and can collect information about a learner using Tobii Pro Studio and Morae in a software configuration.

Specifically, the visual action collecting unit 202 may measure the characteristics of the learner's eyes and calibrate the visual activity data of the learner by allowing the learner to observe the red circle presented by the screen . Thereafter, the visual action collecting unit 202 can track the eyes according to the eyes of the learner who learns the learning video, and can collect visual behaviors through the eyes.

1) EyeTracker uses EyeTracker, which is a non-intrusive measure of learner's visual behavior, to generate data such as time stamp, pupil size (mm), type of eye movement (fixation / saccade) Can be obtained. At this time, EyeTracker can collect visual activity through Tobii Pro Studio and extract the collected data.

2) The webcam can record the learner's learning image by recording the learner watching the learning video in order to conduct the recall stimulus interview. At this time, the webcam records the face of the learner through Morae, and can captain the learner in advance of the section requiring the question.

Here, the recall stimulus interview is a process in which a learner's face on which the learner's face is recorded and the learner's lecture image are provided as stimuli, and the learner's thoughts and feelings It can be a talking interview. At this time, the present invention can mark the point of time according to the learner's movement or change of the facial expression, and then provide the marked image to the learner as a stimulus.

3) The tablet pen can acquire the input information in the display of the user terminal in order to track the learner's gaze.

The visual behavior data generation unit 203 can generate visual behavior data for each learner's visual behavior. The visual behavior data generation unit 203 can classify the visual behaviors of the learner by eye gaze fixation, visual jump, size of the left and right pupils, coordinate values of the eyes, and distance between the eyes and the screen. Here, the coordinate values for the visual line can be calculated as the X and Y coordinate values of the left and right eyes, and the time at which the learner's eyes are located inside the learning region of the learning video. That is, the present invention does not measure coordinate values for the line of sight based on merely the movement of the learner's gaze, but rather measures the learner's gaze in consideration of the time at which the gaze at the position where the learner fixed the gaze is fixed One coordinate value can be measured.

For example, the visual behavior data generation unit 203 can measure the degree of immersion of the learner by comparing the absolute time that the line of sight stays in the learning area. In other words, the longer the learner is immersed in the learning contents, the longer the time for the gaze to stay in the learning video may become longer, and the longer the gaze stay time is, the higher the immersion is in proportion to the time. Therefore, the visual behavior data generation unit 203 can grasp the degree of immersion of the learner using the time during which the visual line in the learning moving picture stays.

Further, the visual behavior data generation unit 203 can not only grasp the linger time of the learner in the learning video, but also check whether the lecturer is particularly emphasized in the entire learning area or whether the visual line is located in the specified area . In detail, the visual behavior data generation unit 203 can determine that the degree of immersion is higher as the degree to which the instructed object and the area in which the visual line stays coincides with the point at which the instructor indicates a part of the on- have.

For example, the visual behavior data generation unit 203 compares the coordinate values of the learner's gaze with the coordinate values at the time point instructed by the instructor to explain the learning contents of the learning video, and the error range between the coordinate values is small It can be judged that the degree of immersion of learners is high.

Then, the visual behavior data generation unit 203 generates a visual behavior data for each of the visual behaviors in accordance with the average eye fixation duration, the average eye jump duration, the average pupil size relative to the baseline, the average eye duration in the learning area, And visual distance data such as a screen distance can be generated.

Here, the average eye fixed duration can be determined by the eye fixation frequency per second / the number of frames per second. The average eye jump duration can be determined by the frequency of gaze hop per second / number of frames per second. The mean pupil size versus baseline can be determined by the mean of the left and right pupil diameter / the mean of the baseline pupil diameter. In this case, in the case of the left and right pupil diameters, it is possible to calculate the change in the size of the pupil relative to the baseline after calculating the baseline by subtracting the pupil size of the individual learner from the normal pupil.

In other words, the present invention can induce a learner to fix the line of sight to the "X" In this case, the present invention can present the symbol "X" once every 10 seconds on the screen, and present a total of 5 times at intervals of 10 seconds. The present invention can minimize the accommodation responses of the eyes by inducing the learner to look at the point of the "X" symbol presented on the screen. The present invention can set the baseline as an average of the pupil diameter for 50 seconds measured based on the thus measured eye coordination reaction.

For example, the present invention relates to a method for adjusting an eye according to a phenomenon of an eye corresponding to a stimulus by an "X" symbol displayed on a screen at a certain point in time before or after learning a moving image, , And can set the baseline according to the adjustment reaction of the eyes.

The average visual line duration in the learning area can be determined by the sum of the number of visual lines located within the learning area per second / the number of frames per second. The eye and the screen distance relative to the pre-learning interval can be determined by the number of frames per second (the average eye and the screen distance between the left and right eyes, the distance between the eyes and the screen before the learning), and the number of frames per second.

The graph generating unit 204 may generate a graph of the occurrence frequency of each learner's visual behavior using the visual behavior data. The graph generating unit 204 may generate a graph corresponding to each of the average values of the duration of the gaze fixation frequency, the duration of the gaze hopping frequency, and the size of the left and right pupils based on the stream of the learning moving image.

The peak interval setting unit 205 can set a peak interval according to the occurrence frequency of the learning moving picture in the entire interval using the generated graph. The peak interval setting unit 205 may set specific time zones within the entire interval based on the graph and may set a point indicating a high occurrence frequency of each of the specific time zones to each peak interval. Here, the specific time zone may be a time zone in which the seconds value of each learner corresponds to more than the ISD + ISD per visual activity.

Specifically, the peak section setting unit 205 can obtain an average and a standard deviation of visual behaviors corresponding to the total learning time for all learners and for all the learners, for all learners who have viewed the learning videos. Here, each learner can have different mean and standard deviation values for each individual. The peak interval setting unit 205 can determine, in seconds, the number of seconds of the learning interval in which the standard deviation value of each student is revealed based on the standard deviation value obtained for each learner, and the individual peak interval (1SD) The time zone in which the above values are shown may be all different.

Thereafter, the peak interval setting unit 205 can analyze the entire peak interval for all learners based on individual peak intervals obtained for each learning. In other words, the peak interval setting unit 205 integrates the individual peak periods of the learners to check the entire peak interval in which individual learner-specific peak intervals occur frequently. . The peak interval setting unit 205 sets the peak time period of the top 10% of the highest number of students who have calculated the peak value as the peak value by summing how many learers have shown the peak value of the learning video in seconds have.

The academic achievement prediction unit 206 can predict the academic achievement level of the learner by analyzing the learner's visual behavior at the set peak interval. In other words, the learning achievement prediction unit 206 can determine information about the learning commitment of the learner with respect to the learning video from the learner's visual behavior. Then, the academic achievement prediction unit 206 can predict the degree of academic achievement through the results of visual behavior and learning commitment.

FIG. 3 is a diagram for explaining learning data constituting a learning moving picture according to an embodiment and visual behavior of a learner according to a behavior of a teacher.

Referring to FIG. 3, the learning achievement prediction device can confirm which learning stimulus in the learning video is responding to the visual behavior data in the learning environment of the learning video. Specifically, the present invention can analyze a lecture video (learning video) used as a stimulus to grasp the characteristics of a visual activity and a cause of a visual activity occurring to a learner while watching a learning video. Accordingly, the present invention can classify the factors that may affect the learner's visual behavior in the lecture image as (1) teaching behavior and (2) learning data.

3 (a) and 3 (b), a learning moving picture may include a teaching behavior 301 and learning data 302 as learning stimuli.

Instructional Behavior (301) is a behavior of instructor who describes learning contents in learning video, and can express different characteristic behaviors according to the object to be explained by the instructor. In other words, the instructor can convey the information to the learner with visual stimuli based on 'objectively identifiable behavior change' in order to explain the learning video to the learner. Accordingly, the present invention can distinguish the teaching behavior 301 from the intuitive operation according to the time stamp of the learning moving picture, which can be expressed as shown in Table 1.

[Table 1]

Table 1 shows the time stamp of the learning video in units of 1 second and expresses the behavior of the instructor in seconds. In addition, the present invention can be expressed as shown in Table 2 after arranging the respective items for the teaching behavior.

[Table 2]

The learning material (302) occupied a certain area in the whole of the learning video, and the type of the learning material is composed of text, graph, and table, and can be utilized as a stimulus to the user.

In other words, the present invention can classify patterns of teaching behaviors constituting learning videos provided to a learner, and map learning materials corresponding to patterns of respective teaching behaviors. In addition, the present invention can extract the visual behavior of a learner at a point of time when learning material mapped in response to a teaching behavior is provided to a learner. Thereafter, the present invention can analyze the degree of the learner's learning commitment to the learning video from the visual activity of the learner at the time point.

3 (c) and 3 (d), the learning achievement prediction device can confirm the link between the visual behavior based on the teaching behavior 301 and the learning data 302. The present invention is based on the assumption that the learner's visual behavior by the instructional action 301 and the learning material 302 after deriving the upper 10% Can be confirmed.

1) Eye fixation during visual action

The gaze fixation is a state in which the gaze is fixed for a certain period of time in the learning video, which may be a state that appears when the learner concentrates on a specific region. For example, gaze fixation is an eye movement that is defined as a gaze time of about 100 to 600 mm. The duration of gaze fixation is the time to process information or perform cognitive activities on an object, Can mean one hour

This gaze fixation can occur frequently in the following sections when compared with teaching behavior. In other words, in the learning environment of the learning video, the teaching behaviors that appear in the section where the gaze fixation frequently occurs can be classified into the following: Occurred frequently. That is, the learner focuses on the visual information (learning material) pointed by the instructor in the course of watching the learning video in real time according to the teaching act, and the present invention maintains the concentrated state corresponding to the gaze fixed duration . At this time, the present invention can measure the coordinate values of the left and right eye lines according to the gaze fixation duration time. Then, it is possible to grasp what the learner observes at the time when the learner fixed the gaze.

2) Visual leap during visual action

Eye-gaze leap refers to a process in which the learner's gaze moves to the interested subject in the learning video and then the gaze is fixed in the moved position. This means that the learner is interested in the learning video, It may be the state that occurred. For example, gaze leap refers to eye gazing at one place and moving quickly to the next destination. An eye movement that shows an average movement of about 600 ° per second with an average of about 20 to 40 ms. This can mean the time at which convergence for concentration occurs.

Such a gaze jump can occur in a region where convergence occurs. That is, in the present invention, when visual leap is frequent, the visual information processing is hardly performed while the visual line leap occurs, and when the instructor indicates a specific part of the learning material and when the screen of the learning material is switched have. In this case, it can be assumed that the instructional behavior of the instructor is commonly matched to the gaze fixation and gaze leap, so that when the instructor takes the instruction action, the learner 's gaze moves to the indication region, stabilizes and leads to gaze fixation .

In addition, it can be seen that when the learning data screen is switched and when the data types presented on the screen are large (text + graph), the visual gaze is frequent. This can be used as a measure to show that learners are concentrating on learning when there is continuous visual leap, visual information processing is not done, have. In other words, it can be seen that the learner responds to the instruction sequence in a sequential manner according to the instructor's order that the gaze jump to a specific part frequently occurs according to the instructional activity. .

3) Changes in pupil size during visual activity

Changes in pupil size are mainly related to cognitive load or emotional changes, which can vary depending on the degree of learner concentration. The present invention can extract only the section where the gaze fixation and gaze jump do not overlap the section where the change of the pupil size is severe.

This is because the pupil size changes when the teacher does unnecessary gestures in the teaching lecture or if the pupil contents are covered by the body and is invisible. This can be confirmed by the external cognitive load, that is, the unnecessary cognitive load caused by the way and manner of presentation of the learning data, or the negative emotional change.

As a result, the present invention can grasp the learner's psychological state through visual behavior, which can be determined by teaching behavior and learning data that cause visual behavior in the learning environment of the learning video.

4 is a diagram for explaining a method of generating a graph of visual behavior according to an embodiment.

Referring to FIG. 4, the learning achievement predicting device generates a graph using the gaze fixed duration, the gaze hop duration, and the average value of the pupil size per second in order to confirm the change of the visual behavior of the learner while listening to the learning video can do.

Referring to the graph of FIG. 4, it can be seen that the gaze fixation has a longer duration in the concept explanatory section than the problem solving section as a whole, and the gaze jump increases finerly over time. In addition, the pupil diameter shows a large value at the beginning of learning, and the pupil diameter tends to decrease as learning progresses. In addition, the pupil diameter is recorded at the lowest value in the last section of the concept explanation, and thereafter, it increases somewhat in the problem solution section.

In order to analyze the characteristics of each visual behavior, the present invention displays the learning data provided in correspondence with the teaching behavior and teaching behavior at the time corresponding to the top 10% of the visual behavior as shown in Table 3 .

[Table 3]

As shown in Table 3, gaze fixation occurs in the order of the instructor (13 sec.), The instructor (8 sec.), The instructor's gaze (4 sec) Can be confirmed. It can be seen that the visual leap occurs in the order of (13 seconds) when instructing a specific part of the learning material by hand (5 seconds), and when writing (4 seconds). Finally, the pupil diameter was found to increase in order of gesture (8 sec.) For the instructor to explain, 7 sec for the gaze movement (6 sec) have.

As a result, gaze fixation duration occurred when the learner had to pay attention to a certain area in the learning video, and gaze jump duration was found to occur mainly when concentration convergence was needed. In addition, the pupil diameter overlapped with the teaching behavior of the upper section where gaze fixation or gaze leap occurred, but there was also a teaching behavior characteristic only in the upper pupil diameter section.

This is the time when the teacher can not confirm the contents of the writing because it takes unnecessary gestures (2 seconds) and the teacher can not confirm the contents of the writing (1 second) . For example, the present invention can confirm that the perceived load is large when the pupil size is enlarged.

[Table 4]

Table 4 shows the result of matching the learner's visual behavior characteristics with the learning data screen types provided in the learning video. The visual fixation and pupil diameter showed the longest duration (25 sec, 44 sec) when the text-oriented data were presented. The gaze leap was longer in the screen including the graph and the text (47 Second). It can be seen from Table 4 that the more the elements presented on the same screen, the more the eye movement becomes.

In the graph-based learning data section, the gaze fixation duration was 18 seconds, 13 seconds of which was when the professor was writing. It can be presumed that the behavior of a changing teacher can be a factor that has a more direct influence on visual behavior than a relatively long fixed learning material,

Through the above results, we could confirm the relationship between visual behavior and teaching behavior, visual behavior and learning materials.

FIG. 5 is a view for explaining a method of setting a peak interval according to the occurrence frequency in the entire section according to an embodiment.

Referring to FIG. 5, the academic achievement prediction apparatus can set a peak period in which a specific visual action frequently occurs in a learning moving picture, and can predict the academic achievement by visual action only in the set peak period. In other words, the learning achievement prediction device can select a peak interval in which a particular visual behavior frequently appears to the entire learner. Here, the peak period can be calculated as the time corresponding to the upper 10% of the frequencies in each of the time zones obtained by calculating the seconds value of each learner in the visual action corresponding to + 1 SD or more over the whole interval.

For example, referring to FIG. 5, the present invention calculates a frequency of occurrence of 1SD or more for each learner in a peak interval of 76 seconds, a visual jump of 96 seconds, and a pupil diameter of 67 seconds, can do.

FIG. 6 is a diagram for explaining a process of predicting an academic achievement according to an embodiment.

Referring to FIG. 6, the learning achievement prediction device may collect visual activity of a learner who watches a learning video, and then generate visual activity data of the learner's visual activity. The learning achievement predicting device can generate a graph of the occurrence frequency of each learner's visual behavior using visual behavior data. After setting a peak period in which a specific visual behavior frequently occurs, the set peak period The value and the frequency for the visual behavior can be calculated by generating the state behavior data for each visual behavior.

The present invention can confirm that the learner's concentration state is maintained when the gaze fixation duration is long in the peak interval, and can be confirmed that the learner's concentration transition is maintained when the duration of the hazard jump is long. In addition, it can be confirmed that the larger the diameter of the pupil size of the present invention is, the more the external perception is generated.

Here, the present invention can predict the academic achievement in consideration of the visual fixed value at the peak interval and the occurrence frequency of 1SD or more of the pupil diameter at the peak of the visual behavior. In this case, the present invention is characterized in that the larger the visual leap value in the peak interval, .

In other words, according to the present invention, as the eye fixed value and the pupil diameter in the sight line section become smaller, it can be understood that the contents of the learner's interest are included in the learning video, and the gaze fixed value and the pupil diameter As a result, it can be an index indicating the concentration of learning by the framework of learning contents and the area of learning contents. Thus, the present invention improves the predictability of learning achievement and predicts achievement with visual behavior in the peak period to derive a reasonable value.

For example, the present invention can predict a student's academic achievement over the entire video based on the results at each peak period. In other words, according to the present invention, in response to the intention of the learner who has taught the learning contents of the learning video and the provider who has produced the learning video, it is understood that the learning progress of the learner is proceeding in the direction intended by the learner, The learner's academic achievement can be predicted.

As another example, the present invention can predict the academic achievement of the topic learner corresponding to each peak period. In other words, the present invention grasps a learner's visual behavior with respect to a peak period in the course of watching a learning video, and can grasp a learner's competence such as a concept description corresponding to a small classification, have. In addition, the present invention further analyzes the visual behavior as a peak section of each learning section and analyzes the visual behavior, and it can be understood that the concentration of the concept is more improved at the point of the concept of the structure explained. That is, according to the present invention, by setting a peak interval at intervals of 1 second, the learning contents of the learning video can be grasped at the peak interval corresponding to the top 10% of the peak interval, It is possible to predict academic achievement.

Based on this, the present invention can be appropriately utilized in designing lecture images to improve learner's academic achievement. In other words, the present invention can constitute a learning video in which a teacher takes a gesture or writes in a written form in a context in which a learner's concentration is particularly required. Further, according to the present invention, it is possible to improve the quality of the learning video and improve the academic achievement by providing additional data or improving the teaching method according to the expected learner's psychological state of the prepared learning video .

Further, the present invention sets desirable visual behaviors to be expected in a specific time and specific area in a learning moving picture, and can use the learning process of the learner as evaluation and achievement prediction on the basis thereof. In other words, the present invention can utilize the learning process for evaluation and achievement prediction by monitoring how much the learner has fixed his / her gaze on the part where concentration is required, and how the instructional action of the instructor agrees with the gaze leap.

The present invention grasps the learning achievement by the peak period of the learners based on the pre-produced learning video, thereby grasping the problem in the learning video for the learners whose learning progress does not progress according to the intention of the producer or the leader, You can also modify your learning videos to drive additional learning instructional methods or to increase your concentration through attentional ventilation.

Further, the present invention can improve the learning effect through the evaluation problem according to the learning concentration of the learner by adjusting the difficulty of the learning video according to the learning concentration of the learner by grasping the learning concentration and the learning achievement according to the system of the learning video .

7 is a flowchart for explaining a method of predicting an academic achievement according to an embodiment.

In step 701, the learning achievement prediction device may collect visual behaviors of the learner who watches the learning video. Here, the learning achievement predicting device can collect visual behaviors regarding the movement of the eyes of the learner watching the learning video, the change of the pupil, and the like.

In step 702, the learning achievement prediction device may generate visual behavior data for each learner's visual behavior. Specifically, the learning achievement prediction apparatus can utilize the visual behavior of the learner responding to at least one of the instructional instruction of the instructor in the lecture screen or the learning material provided according to the time stream of the learning video.

2) the pupil size based on the pupil diameter in the average learning segment of the learner; and 3) the size of the left and right pupils based on the pupil diameter in the average learning segment of the learner. 3) A coordinate value at the time of the lecture screen in the lecture screen of the learning video, and 4) a distance between the learner's eye considering the change of the perspective and the screen on which the learning video is displayed. At this time, the size of the left and right pupils can be determined by the presentation style of the learning data provided through the learning animation and the cognitive load or emotion change of the learner caused by the presentation method.

In step 703, the learning achievement prediction device can generate a graph of the occurrence frequency of the learner's visual behavior using visual behavior data. In detail, the learning achievement prediction apparatus can generate a graph corresponding to each of the average values of the duration of the gaze fixation frequency, the duration of the gaze hopping frequency, and the size of the left and right pupils based on the lecture flow of the learning moving image.

In step 704, the academic achievement prediction apparatus can set a peak period according to the occurrence frequency in the entire section of the learning moving image using the generated graph. Here, the academic achievement prediction apparatus may set specific time zones within the entire interval based on a graph, and may set a point indicating a high occurrence frequency of each of the specific time zones to each peak interval.

In step 705, the learning achievement prediction apparatus can predict the degree of achievement of the learner by analyzing the learner's visual behavior in the set peak period.

Meanwhile, the method according to the present invention may be embodied as a program that can be executed by a computer, and may be embodied as various recording media such as a magnetic storage medium, an optical reading medium, and a digital storage medium.

Implementations of the various techniques described herein may be implemented in digital electronic circuitry, or in computer hardware, firmware, software, or combinations thereof. Implementations may be implemented in a computer program product, such as an information carrier, e.g., a machine readable storage device, such as a computer readable storage medium, for example, for processing by a data processing apparatus, Apparatus (computer readable medium) or as a computer program tangibly embodied in a propagation signal. A computer program, such as the computer program (s) described above, may be written in any form of programming language, including compiled or interpreted languages, and may be stored as a stand-alone program or in a module, component, subroutine, As other units suitable for use in the present invention. A computer program may be deployed to be processed on one computer or multiple computers at one site or distributed across multiple sites and interconnected by a communications network.

Processors suitable for processing a computer program include, by way of example, both general purpose and special purpose microprocessors, and any one or more processors of any kind of digital computer. Generally, a processor will receive instructions and data from a read-only memory or a random access memory or both. The elements of a computer may include at least one processor for executing instructions and one or more memory devices for storing instructions and data. Generally, a computer may include one or more mass storage devices for storing data, such as magnetic, magneto-optical disks, or optical disks, or may receive data from them, transmit data to them, . ≪ / RTI > Information carriers suitable for embodying computer program instructions and data include, for example, semiconductor memory devices, for example, magnetic media such as hard disks, floppy disks and magnetic tape, compact disk read only memory A magneto-optical medium such as a floppy disk, an optical disk such as a DVD (Digital Video Disk), a ROM (Read Only Memory), a RAM , Random Access Memory), a flash memory, an EPROM (Erasable Programmable ROM), an EEPROM (Electrically Erasable Programmable ROM), and the like. The processor and memory may be supplemented or included by special purpose logic circuitry.

In addition, the computer-readable medium can be any available media that can be accessed by a computer, and can include both computer storage media and transmission media.

While the specification contains a number of specific implementation details, it should be understood that they are not to be construed as limitations on the scope of any invention or claim, but rather on the description of features that may be specific to a particular embodiment of a particular invention Should be understood. Certain features described herein in the context of separate embodiments may be implemented in combination in a single embodiment. Conversely, various features described in the context of a single embodiment may also be implemented in multiple embodiments, either individually or in any suitable subcombination. Further, although the features may operate in a particular combination and may be initially described as so claimed, one or more features from the claimed combination may in some cases be excluded from the combination, Or a variant of a subcombination.

Likewise, although the operations are depicted in the drawings in a particular order, it should be understood that such operations must be performed in that particular order or sequential order shown to achieve the desired result, or that all illustrated operations should be performed. In certain cases, multitasking and parallel processing may be advantageous. Also, the separation of the various device components of the above-described embodiments should not be understood as requiring such separation in all embodiments, and the described program components and devices will generally be integrated together into a single software product or packaged into multiple software products It should be understood.

It should be noted that the embodiments of the present invention disclosed in the present specification and drawings are only illustrative of specific examples for the purpose of understanding and are not intended to limit the scope of the present invention. It will be apparent to those skilled in the art that other modifications based on the technical idea of the present invention are possible in addition to the embodiments disclosed herein.

101: How to predict academic achievement.
102: user terminal
103; Learning videos
104: Instructor
105: Learning Resources
106: Learners
107: Visual behavior

Claims (12)

Collecting a visual activity of a learner watching a learning video;
Generating visual activity data for each of the learner's visual behaviors;
Generating a graph of an occurrence frequency of each learner's visual behavior using the visual behavior data;
Setting a peak interval according to an occurrence frequency in the entire section of the learning moving picture using the generated graph; And
Analyzing a visual activity of a learner in the set peak period and predicting a degree of academic achievement of the learner;
A method for predicting academic achievement. The method according to claim 1,
Wherein the step of generating the visual behavior data comprises:
And a visual behavior of the learner responding to at least one of the instructional instruction of the instructor in the lecture screen or the learning material provided according to the time flow of the learning video. The method according to claim 1,
Wherein the step of generating the visual behavior data comprises:
2) the size of the left and right pupils based on the pupil diameter in the average learning segment of the learner; and 3) the learning on the screen of the lecture screen of the learning video. A coordinate value at the time the gaze is located, and 4) a distance between the learner's eye considering the change of perspective and the screen on which the learning video is displayed. The method of claim 3,
The size of the left and right pupils is,
The learning achievement prediction method determined by the learner's cognitive load or emotional change caused by the presentation style and the presentation style of the learning material provided through the learning video The method of claim 3,
Wherein generating the graph comprises:
And generating a graph corresponding to each of the average values of the duration of the gaze fixation frequency, the duration of the gaze hopping frequency, and the size of the left and right pupils based on the stream of the learning moving image. The method according to claim 1,
Wherein the setting of the peak interval comprises:
And setting a specific time zone in the whole section based on the graph and setting a point showing a high occurrence frequency of each visual behavior in each of the specific time zones to each peak period. In the academic achievement prediction system,
A processor,
The processor comprising:
The visual activity of the learner who watches the learning video is collected,
Generates visual behavior data for each learner's visual behavior,
Generating a graph of occurrence frequency of each learner's visual behavior using the visual behavior data,
Setting a peak interval according to occurrence frequency in the entire section of the learning moving picture using the generated graph,
And analyzing a learner's visual behavior in the set peak period to predict the academic achievement level of the learner. 8. The method of claim 7,
The processor comprising:
And a visual behavior of the learner responding to at least one of the instructional instruction of the instructor in the lecture screen or the learning material provided according to the time flow of the learning video. 8. The method of claim 7,
The processor comprising:
2) the size of the left and right pupils based on the pupil diameter in the average learning segment of the learner; and 3) the learning on the screen of the lecture screen of the learning video. A coordinate value at the time the gaze is located, and 4) a distance between the eye of the learner considering the change of the perspective and the screen on which the learning video is displayed. 10. The method of claim 9,
The size of the left and right pupils is,
The learning achievement prediction system determined by the learner's cognitive load or emotional change caused by the presentation form and presentation method of the learning material provided through the learning video. 10. The method of claim 9,
The processor comprising:
And generates a graph corresponding to each of the average values of the duration of the gaze fixation frequency, the duration of the gaze hopping frequency, and the size of the left and right pupils based on the lecture flow of the learning moving picture. 8. The method of claim 7,
The processor comprising:
Setting a specific time zone within the entire section based on the graph and setting a point showing a high occurrence frequency of each visual behavior in each of the specific time zones to each peak section.

Images