Teaching material provision device, teaching material provision method, and program
The teaching material providing device analyzes learner data to prioritize content based on overall and individual learning trends, addressing the challenge of unclear prioritization in conventional technologies by providing personalized materials that align with student needs, thereby improving learning efficiency.
Patent Information
- Application Number
- PCT/JP2023/019822
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2023-05-26
- Publication Date
- 2024-12-05
AI Technical Summary
Conventional personalized learning technologies fail to accurately identify which pages are most important for students to study, as they do not adequately consider individual learning tendencies and characteristics, leading to unclear prioritization of study materials.
A teaching material providing device and method that analyzes teaching materials and learner data to derive importance levels based on overall and individual learning tendencies, using a performance prediction model to calculate weights for feature quantities such as reading behavior and quiz results, and provides personalized materials to students.
Effectively provides personalized teaching materials to students by accurately identifying and prioritizing important content based on individual and collective learning trends, enhancing learning efficiency and understanding.
Smart Images

Figure JP2023019822_05122024_PF_FP_ABST
Abstract
Description
Educational material providing device, educational material providing method, and program
[0001] The disclosed technology relates to a teaching material providing device, a teaching material providing method, and a program.
[0002] Recently, there has been an increasing trend to use learning materials that are personalized for each learner in order to achieve effective learning. However, creating such learning materials requires not only extracting important points based on the words, sentences, images, etc. contained in the materials, but also taking into consideration the learning tendencies of each learner (level of comprehension, whether it is easier to improve grades by retaking quizzes than by reading the materials, etc.) as well as the learning tendencies of all learners using the materials (such as the parts they focus on reading, etc.).
[0003] As a technology for improving the efficiency of review by providing personalized learning points to target students, for example, the technology described in Non-Patent Document 1 is disclosed. The technology described in Non-Patent Document 1 provides target students with pages related to quizzes that they got wrong, pages on which they and many other students pressed the "I didn't understand" button, a graph of the viewing time for each page for all students, and links to the relevant pages.
[0004] Fumiya Okubo, Tetsuya Shiino, Tsubasa Minematsu, Yuta Taniguchi, Atsushi Shimada Adaptive Learning Support System Based on Automatic Recommendation of Personalized Review Materials IEEE Transactions on Learning Technologies, 2022.11.
[0005] The above-mentioned conventional technology has the problem that it is difficult for the target student to know which pages to prioritize for study because it provides pages to review from multiple perspectives. Furthermore, while the conventional technology takes into account the learning results of the target student, it does not take into account the learning characteristics (study methods, etc.) or the characteristics of the subject (what should be studied with emphasis to improve grades, etc.), so there is a problem in that it is unclear whether the pages provided are truly important to the target student.
[0006] The technology disclosed herein has been made in consideration of the above points, and provides a teaching material providing device, a teaching material providing method, and a program that can provide personalized teaching materials to target learners.
[0007] A first aspect of the present disclosure is a teaching material providing device comprising: a teaching material analysis unit that analyzes teaching materials and derives a teaching material-specific importance for each range of the teaching materials; an overall learning tendency analysis unit that analyzes the learning tendencies of multiple learners who are using the teaching materials together and derives a level of importance based on the overall learning tendency for each range of the teaching materials; an individual learning tendency analysis unit that analyzes the learning tendencies of individual learners to whom the teaching materials are to be provided and derives a level of importance based on the individual learning tendency for each range of the teaching materials; an importance calculation unit that calculates an importance for each range of the teaching materials using the teaching material-specific importance, the importance based on the overall learning tendency, the importance based on the individual learning tendency, and weights of the features obtained by a grade prediction model in which learning behavior or learning results are used as features; and a teaching material providing unit that provides the range of teaching materials extracted from the teaching materials based on the importance to the learners to whom the teaching materials are to be provided.
[0008] A second aspect of the present disclosure is a teaching material provision device in which, in the teaching material provision device of the first aspect, the overall learning trend analysis unit derives, for each feature, an importance in the overall learning trend related to the feature, the individual learning trend analysis unit derives, for each feature, an importance in the individual learning trend related to the feature, and the importance calculation unit multiplies, for each feature, the importance based on the overall learning trend related to the feature, the importance based on the individual learning trend related to the feature, and the weight of the feature to calculate a multiplied value, and adds the importance specific to the teaching material and the multiplied value calculated for each feature to calculate the importance of the range.
[0009] A third aspect of the present disclosure provides the learning material providing device according to the first aspect, wherein the weight is a contribution of the feature amount obtained by the grade prediction model.
[0010] Furthermore, a fourth aspect of the present disclosure is a method for providing teaching materials, which analyzes teaching materials and derives a teaching material-specific importance for each range of the teaching materials, analyzes the learning tendencies of multiple learners who are using the teaching materials as a whole and derives a level of importance based on the overall learning tendency for each range of the teaching materials, analyzes the learning tendencies of individual learners to whom the teaching materials are to be provided and derives a level of importance based on the individual learning tendency for each range of the teaching materials, calculates an importance for each range of the teaching materials using the teaching material-specific importance, the importance based on the overall learning tendency, the importance based on the individual learning tendency, and weights of the features obtained by a grade prediction model that uses learning behavior or learning results as features, and provides the range of teaching materials extracted from the teaching materials based on the importance to the learners to whom the teaching materials are to be provided.
[0011] A fifth aspect of the present disclosure is a program for causing a computer to function as each unit of the educational material providing device according to any one of the first to third aspects.
[0012] The techniques disclosed herein can provide personalized learning materials to target learners.
[0013] FIG. 1 is a diagram for explaining a teaching material providing device of an embodiment. FIG. 1 is a configuration diagram showing an example of a hardware configuration of a teaching material providing device of an embodiment. FIG. 2 is a functional block diagram showing an example of a configuration of a teaching material providing device in carrying out a weight determination process. FIG. 3 is a flowchart showing an example of a weight determination process executed by the teaching material providing device of an embodiment. FIG. 4 is a functional block diagram showing an example of a configuration of a teaching material providing device in carrying out a teaching material providing process. FIG. 5 is a flowchart showing an example of a teaching material providing process executed by the teaching material providing device of an embodiment. FIG. 6 is a diagram for explaining an example. FIG. 7 is a diagram for explaining an example. FIG. 8 is a diagram for explaining an example. FIG. 9 is a diagram for explaining an example. FIG. 10 is a diagram for explaining an example.
[0014] Hereinafter, an example of an embodiment of the disclosed technology will be described with reference to the drawings. Note that the same or equivalent components and parts in each drawing are given the same reference numerals.
[0015] The teaching material providing device of this embodiment is a device that, in a situation where multiple students use teaching materials, provides each student with teaching materials that are personalized for that student from a master teaching material. Note that the master teaching material may be one teaching material or multiple teaching materials. For example, when multiple teaching materials are used for a certain subject, the teaching material providing device may be a device that provides personalized teaching materials for each student using the multiple teaching materials being used as master teaching materials. In this case, the teaching material providing device can also merge the multiple teaching materials and process them as a single teaching material.
[0016] 1, the teaching material providing device 20 provides each of n students, Student 1 to Student n, with teaching materials 11_1 to 11_s, which are personalized versions of the teaching material 10. For example, when the teaching material 10 includes multiple pages, the teaching material providing device 20 provides a personalized page from among the multiple pages to the target student as the teaching material. Note that the students in this embodiment are an example of learners in the present disclosure.
[0017] 2 is a configuration diagram showing an example of the hardware configuration of the learning material providing device 20 of this embodiment. As shown in Fig. 2, the learning material providing device 20 includes a CPU (Central Processing Unit) 51, a ROM (Read Only Memory) 52, a RAM (Random Access Memory) 53, a storage 54, a display unit 57, and a communication I / F (Interface) 58. Each component is connected to each other so as to be able to communicate with each other via a bus 59 such as a system bus or a control bus.
[0018] The CPU 51 is a central processing unit, and executes various programs such as a weight determination program 55A and a teaching material providing program 55B stored in the storage 54, and controls each section.
[0019] The ROM 52 stores various data to be executed by the CPU 51. The RAM 53 temporarily stores programs or data as a work area when the CPU 51 executes various programs. That is, the CPU 51 reads a program from the storage 54 and executes the program using the RAM 53 as a work area.
[0020] In this embodiment, the storage 54 stores a weight determination program 55A and a learning material provision program 55B. Each of the weight determination program 55A and the learning material provision program 55B may be a single program or a group of programs consisting of multiple programs or modules. The storage 54 is configured with an HDD (Hard Disk Drive) or an SSD (Solid State Drive). The storage 54 also stores various programs, including an operating system, and various data (neither of which are shown). The storage 54 also stores a performance prediction model 56A and weight information 56B, which will be described in detail below.
[0021] The display unit 57 displays various information such as information relating to the provision of personalized teaching materials, etc. The display unit 57 is not particularly limited, and may be any of various displays.
[0022] The communication I / F 58 is an interface for communicating with external devices, and uses standards such as Ethernet (registered trademark), FDDI, and Wi-Fi (registered trademark).
[0023] The above-described learning material providing device 20 derives the importance for each student for each range of the learning material 10 using the importance specific to the learning material, the importance based on the learning tendencies of all the students using the learning material 10, the importance based on the learning tendencies of each individual student, and the weight of the feature obtained by the grade prediction model 56A, which uses learning behavior or learning results as feature quantities.The learning material providing device 20 then provides personalized learning materials 11 to each student based on the importance for each range of the learning material 10.
[0024] Hereinafter, the process of providing personalized learning materials 11 to each student in the learning material providing device 20 will be described with reference to the drawings.
[0025] [Weight Determination Process] First, the learning material providing device 20 performs a process of determining the above-mentioned "weight" as a preparation. First, the "weight" will be explained. The weight is a weight for adjusting the degree of influence on the importance based on the learning tendency. First, the learning material providing device 20 of this embodiment performs a process of determining the weight as a preparation.
[0026] 3 is a functional block diagram showing an example of the configuration of the learning material providing device 20 when performing the weight determination process. As shown in FIG. 3, the learning material providing device 20 includes a learning material pre-processing unit 22, an overall learning pre-processing unit 24, an individual learning pre-processing unit 26, a course information pre-processing unit 28, a grade prediction model building unit 30, and a weight determination unit 32. In this embodiment, the CPU 51 executes the weight determination program 55A, causing the CPU 51 to function as the learning material pre-processing unit 22, the overall learning pre-processing unit 24, the individual learning pre-processing unit 26, the course information pre-processing unit 28, the grade prediction model building unit 30, and the weight determination unit 32.
[0027] The learning material pre-processing unit 22 performs predetermined pre-processing on the learning material data of the learning material 10. In this embodiment, as an example, the learning material 10 is stored in a data server device (not shown) external to the learning material providing device 20, and the learning material data of the learning material 10 is input from the external data server device. Examples of the learning material data of the learning material 10 include PowerPoint data, document data, still image data, and moving image data, but are not limited thereto and may be data in any unit. Examples of pre-processing include normalization of text data, removal of HTML tags, tokenization, removal of stop words, and vectorization, but are not limited thereto and may be any processing appropriate for subsequent processing. The learning material pre-processing unit 22 outputs the pre-processed learning material data to the grade prediction model building unit 30.
[0028] The overall learning preprocessing unit 24 performs predetermined preprocessing on the overall learning log 12. In this embodiment, as an example, the overall learning log 12 is stored in a data server device (not shown) external to the learning material providing device 20, and the overall learning log 12 is input from the external data server device. The overall learning log 12 is a record of learning using the learning material 10 by all students using the learning material 10, and includes, for example, data on the results of learning (hereinafter referred to as "learning result data") and data on viewing of the learning material (hereinafter referred to as "learning material viewing data"). Examples of learning result data include quiz results, assignment results, and final grades, but are not limited to these and may be any data. Examples of learning material viewing data include the time spent viewing the learning material and the locations of highlights on the learning material, but are not limited to these and may be any data. The preprocessing involves converting the overall learning log 12 into data on the learning trends of all students using the learning material 10 (hereinafter referred to as "overall learning trends"). Examples of preprocessing include, but are not limited to, conversion into a predetermined format such as numerical values, character strings, date / time data, etc. Regarding learning results, examples of overall learning trends include the accuracy rate of tests and assignments, average scores, and score distribution, but are not limited to these and may be any data. Regarding viewing of learning materials, examples include the number of words highlighted, the average viewing time per page of the learning materials, and the page review rate, but are not limited to these and may be any data. The overall learning preprocessing unit 24 outputs the overall learning trends obtained by preprocessing to the grade prediction model construction unit 30.
[0029] The individual learning preprocessing unit 26 performs predetermined preprocessing on the individual learning log 14. In this embodiment, as an example, the individual learning log 14 is stored in a data server device (not shown) external to the learning material providing device 20, and the individual learning log 14 is input from the external data server device. The individual learning log 14 is a record of the learning of the individual student using the learning materials 10 to which the personalized learning materials 11 are provided, and includes, for example, learning result data and learning material viewing data. Examples of learning result data include quiz results, assignment results, and final grades, but are not limited thereto and may be any data. Examples of learning material viewing data include the time spent viewing the learning materials and the locations of highlights in the learning materials, but are not limited thereto and may be any data. The preprocessing involves converting the individual learning log 14 into learning tendency data of the individual student (hereinafter referred to as "individual learning tendency"). Examples of preprocessing include, but are not limited to, conversion into a predetermined format such as a numerical value, a character string, or date / time data. Incidentally, examples of individual learning tendencies include test and assignment scores and response times for learning results, but are not limited to these and may be any data. Also, examples of individual learning tendencies include the number of highlights placed, the viewing time per page of the learning material, and the page review rate, but are not limited to these and may be any data. The individual learning preprocessing unit 26 outputs the individual learning tendencies obtained by preprocessing to the grade prediction model construction unit 30.
[0030] The course information preprocessing unit 28 performs predetermined preprocessing on the course information 16. In this embodiment, as an example, the course information 16 is stored in a data server device (not shown) external to the teaching material providing device 20, and the course information 16 is input from the external data server device. The course information 16 is information related to learning / lessons using the teaching material 10, such as a syllabus, but is not limited to this and may be any data. The preprocessing is a process of converting the course information 16 into data representing the trend of the course (hereinafter referred to as the "course trend"). Examples of preprocessing include, but are not limited to, clusters to which the course belongs, classified based on evaluation criteria such as the allocation of points for quizzes and regular tests. The course information preprocessing unit 28 outputs the course trend obtained by the preprocessing to the grade prediction model construction unit 30.
[0031] The grade prediction model construction unit 30 creates features representing learning behavior or learning results from the preprocessed learning material data, overall learning tendencies, individual learning tendencies, and course tendencies, and trains a machine learning model using the created features as input and the grade prediction results as output to construct grade prediction model 56A. Specifically, multiple training data are prepared using combinations of features and grades as training data, and grade prediction model 56A is trained, including the contribution of each feature to grade. The grade prediction model construction unit 30 stores the constructed grade prediction model 56A in storage 54.
[0032] The weight determination unit 32 uses the grade prediction model 56A to determine weights for adjusting the influence of the feature on the importance based on learning tendencies. In this embodiment, the contribution of a feature included in the grade prediction model 56A is used as the weight for that feature. Note that methods used by the weight determination unit 32 to derive the contribution include techniques such as LIME (Local Interpretable Model-agnostic Explanations), but are not particularly limited, and any known technique can be used. As an example, the contribution in this embodiment is expressed as a numerical value ranging from 0 to 1, with "1" representing the highest contribution and "0" representing the lowest contribution. The weight determination unit 32 stores information representing the determined weight in the storage 54 as weight information 56B.
[0033] Fig. 4 shows a flowchart of an example of weight determination processing executed by the learning material providing device 20 of this embodiment. The learning material providing device 20 executes the weight determination processing shown in Fig. 4 by having the CPU 51 execute a weight determination program 55A stored in the storage 54. Note that the timing for executing the weight determination processing shown in Fig. 4 is not particularly limited, and it may be any timing before personalized learning material is generated.
[0034] In step S10 of FIG. 4, the learning material pre-processing unit 22 pre-processes the learning material data of the learning material 10 as described above, and outputs the pre-processed learning material data to the grade prediction model building unit 30.
[0035] In the next step S12 , the overall learning pre-processing unit 24 performs pre-processing on the overall learning log 12 as described above, and outputs the overall learning tendency to the grade prediction model building unit 30 .
[0036] In the next step S14, the individual learning preprocessing unit 26 performs preprocessing on the individual learning log 14 as described above, and outputs the individual learning tendency to the grade prediction model building unit 30.
[0037] In the next step S16 , the course information preprocessing section 28 preprocesses the course information 16 as described above, and outputs the course tendency to the result prediction model building section 30 .
[0038] In the next step S18, the grade prediction model construction unit 30 performs a grade prediction model construction process. As described above, the grade prediction model construction unit 30 creates feature quantities representing learning behavior or learning results from the overall learning tendency, individual learning tendency, and course tendency. The grade prediction model construction unit 30 then trains a machine learning model using the created feature quantities as input and the grade prediction results as output, constructs grade prediction model 56A, and stores it in storage 54.
[0039] In the next step S20, the weight determination unit 32 performs a process of determining weights. As described above, the weight determination unit 32 derives the contribution of each feature amount to the performance prediction model 56A, determines the weight for that feature amount, and stores weight information 56B in the storage 54. When the process of step S20 is completed, the weight determination process shown in FIG. 4 is completed, and the weight information 56B is prepared.
[0040] [Learning Material Providing Process] Next, a learning material providing process executed in the learning material providing device 20 for providing personalized learning materials 11 to each student will be described.
[0041] 5 is a functional block diagram showing an example of the configuration of the learning material providing device 20 when carrying out the learning material providing process. As shown in FIG. 5, the learning material providing device 20 includes a learning material pre-processing unit 22, a learning material analysis unit 23, an overall learning pre-processing unit 24, an overall learning trend analysis unit 25, an individual learning pre-processing unit 26, an individual learning trend analysis unit 27, an importance calculation unit 36, and a learning material providing unit 38. In this embodiment, the CPU 51 executes the learning material providing program 55B, causing the CPU 51 to function as the learning material pre-processing unit 22, the learning material analysis unit 23, the overall learning pre-processing unit 24, the overall learning trend analysis unit 25, the individual learning pre-processing unit 26, the individual learning trend analysis unit 27, the importance calculation unit 36, and the learning material providing unit 38.
[0042] The learning material pre-processing unit 22, the overall learning pre-processing unit 24, and the individual learning pre-processing unit 26 are the same as those described with reference to FIG. 3 in the weight determination process, and therefore a description thereof will be omitted.
[0043] The learning material analysis unit 23 performs a predetermined analysis on the preprocessed learning material data of the learning material 10 input from the learning material preprocessing unit 22, and derives a learning material-specific importance for each predetermined range. For example, the learning material analysis unit 23 derives the learning material-specific importance from the importance of words in the learning material, the importance of pages in the learning material, and the importance of each time period in the video. An example of a predetermined range is each page. The learning material analysis unit 23 outputs the derived learning material-specific importance for each predetermined range to the importance calculation unit 36.
[0044] The overall learning trend analysis unit 25 performs a predetermined analysis of the overall learning trends input from the overall learning preprocessing unit 24 and derives an importance level based on the overall learning trend for each predetermined range. For example, the overall learning trend analysis unit 25 derives an importance level based on the overall learning trend based on criteria such as a page viewed by many students being highly important, a word highlighted by many students being highly important, and a question in a quiz that many students answered correctly being highly important. The overall learning trend analysis unit 25 outputs the importance level for each predetermined range based on the derived overall learning trend to the importance calculation unit 36.
[0045] The individual learning tendency analysis unit 27 performs a predetermined analysis on the individual learning tendency input from the individual learning preprocessing unit 26 and derives an importance level based on the individual learning tendency for each predetermined range. For example, the individual learning tendency analysis unit 27 derives an importance level based on the individual learning tendency based on criteria such as: a page related to a question that was answered incorrectly in a quiz is highly important; a word that the individual has not highlighted but that many students have highlighted is highly important; and a page that the individual has not viewed but that many students have viewed is highly important. The individual learning tendency analysis unit 27 outputs the importance level for each predetermined range based on the derived individual learning tendency to the importance calculation unit 36.
[0046] The importance calculation unit 36 calculates the importance of each student for each predetermined range of the learning material 10 using the importance specific to the learning material, the importance based on the overall learning tendency, the importance based on the individual learning tendency, and the weight information 56B. If the learner is represented by id, the learning material is represented by material, and the range is represented by range, the importance I(material, range, id) of the learner with respect to the range of the learning material can be calculated by the following formula (1).
[0047]
[0048] In the above formula (1), "I content (material, range)" represents the inherent importance of each range of content-based teaching materials. "K" is a set of features that represent learning behavior or learning results. For example, K = {read, marker, test, subject, ...}. "ω" represents the importance of each range of content-based teaching materials. k " is a weight that indicates the influence on the importance based on the learning tendency of the feature amount k, and corresponds to the weight information 56B. k (material, range)" represents the importance based on the overall learning tendency related to the feature k. k (material, range, id) represents the importance based on the personal learning tendency related to the feature k.
[0049] The importance calculation unit 36 outputs the calculated importance of the teaching material within a predetermined range for each student to the teaching material provision unit 38 .
[0050] The teaching material providing unit 38 extracts a predetermined number of predetermined ranges from the teaching material 10 for each student in descending order of importance based on the importance of each predetermined range of the teaching material for each student, and provides the extracted predetermined number of predetermined ranges to the student as personalized teaching material 11. For example, if the importance of each page of the teaching material 10 is calculated, the top seven pages with the highest importance are provided to the student as personalized teaching material 11.
[0051] Fig. 6 shows a flowchart of an example of the teaching material providing process executed by the teaching material providing device 20 of this embodiment. The teaching material providing device 20 executes the teaching material providing process shown in Fig. 6 by having the CPU 51 execute the teaching material providing program 55B stored in the storage 54. The teaching material providing program 55B of this embodiment is an example of a program disclosed herein. The timing of performing the teaching material providing process shown in Fig. 6 is not particularly limited, and may be any timing, such as a timing instructed by the user (the provider of the personalized teaching material 11).
[0052] The processes in steps S100 to S104 in FIG. 6 are the same as steps S10 to S14 in the weight determination process (see FIG. 4), and therefore a description thereof will be omitted.
[0053] In the next step S106, the learning material analysis section 23 derives the importance specific to the learning material for each predetermined range, as described above.
[0054] In the next step S108, the overall learning tendency analysis unit 25 derives the importance based on the overall learning tendency as described above. Specifically, for one feature value k in the set K, "I k (material, range)" is derived.
[0055] In the next step S110, the individual learning tendency analysis unit 27 derives the importance based on the individual learning tendency as described above. Specifically, for one feature value k in the set K, "P k (material, range, id)" is derived.
[0056] In the next step S112, the importance calculation unit 36 determines whether the above-mentioned set K has become empty. That is, it determines whether the importance has been derived for all feature quantities k included in set K. Until set K becomes empty, the determination in step S112 becomes negative, and the process returns to step S108, and the processes of steps S108 and S110 are repeated.
[0057] On the other hand, if the set K becomes empty, the determination in step S112 becomes positive, and the process proceeds to step S114. In step S114, the importance calculation unit 36 calculates the importance I(material, range, id) for each range of the teaching material using the above formula (1), as described above.
[0058] In the next step S116, as described above, the teaching material providing unit 38 extracts a predetermined number of ranges for each student in descending order of importance I (material, range, id) for each range of the teaching material, and generates personalized teaching material 11 for that student. The teaching material providing unit 38 provides the generated personalized teaching material 11 to the target student. When the processing of step S116 ends, the teaching material providing process shown in FIG. 6 ends.
[0059] (Example) An example will be described. In this example, the teaching material 10 is lecture material such as a textbook or presentation material, the learner who provides the personalized teaching material 11 is learner A, and the predetermined range from which importance is derived is the page of the teaching material 10. In this case, the above formula (1) becomes the following formula (2). In the following formula (2), "I content (lecture material, page)" indicates the importance of each page of the lecture material.
[0060]
[0061] Below, we will explain in detail the case where the feature k "read" represents the viewing of the conference materials during the lecture, the feature k "test" represents a quiz during the lecture, and the set of features K = {read, test}. In this case, the above formula (2) further becomes the following formula (3).
[0062]
[0063] First, the first term on the right side of the above equation (3) is "I contentThe calculation of "(lecture material, page)" will be described. Regarding the text portion of the lecture material, the teaching material preprocessing unit 22 performs text analysis on the lecture material to extract text and words (nouns) containing technical terms from the lecture material. Next, as shown in FIG. 7A, the teaching material analysis unit 23 calculates the TF-IDF (Term Frequency-Inverse Document Frequency) value of words on each page, and then determines the importance of each word by taking the average TF-IDF value across all pages. The sum of the importance of words appearing on each page is normalized to determine the importance of each page. In the example shown in FIG. 7A, the importance of each of the words "basic," "data analysis," "procedure," "business," and "problem definition" among the words extracted from the lecture material is shown. Furthermore, the teaching material analysis unit 23 determines the importance of each page by dividing the sum of the word importance by the maximum value and normalizing it. In the example shown in FIG. 7A, the importance of page 12 is "0.67" and the importance of page 13 is "1.00."
[0064] Furthermore, for the image portion of the lecture materials, the teaching material pre-processing unit 22 performs image analysis to calculate and normalize the difference in image information amount between the previous and next pages. The smaller the difference in image information amount, i.e., the more similar the images are to the previous and next pages, the lower the importance is determined to be. The larger the normalized value is used. The teaching material analysis unit 23 determines the importance of each page by dividing the difference in image information amount by the maximum value and normalizing it. In the example shown in FIG. 7B, the importance of page 12 is "0.30," the importance of page 13 is "0.50," and the importance of page 14 is "0.20."
[0065] The final importance of each page is derived from the importance of each page based on the text analysis and the importance of each page based on the image analysis. In this embodiment, as shown in FIG. 7C, the learning material analysis unit 23 derives the product of the importance of each page based on the text analysis (see FIG. 7A) and the importance of each page based on the image analysis (see FIG. 7B). Note that instead of the product, a weighted sum may be applied to derive the sum. Furthermore, as shown in FIG. 7C, the learning material analysis unit 23 calculates the importance I of each page by dividing the product of the importance by the maximum value and normalizing it. contentDerive (lecture materials, page).
[0066] Next, in the second term on the right side of the above equation (3), "I read (Lecture materials, page) P read As shown in Figures 8A and 8B, the following calculations are performed for (lecture material, page, learner A's ID) using 1) to 5).
[0067] 1) The overall learning tendency analysis unit 25 calculates the number of page views of the lecture materials for all learners attending the lecture using the lecture materials, and read (Lecture materials, page)
[0068] 2) The individual learning tendency analysis unit 27 calculates the median viewing time of all learners divided by the viewing time of learner A.
[0069] 3) The individual learning tendency analysis unit 27 determines which cluster Learner A belongs to using each learner's lecture material viewing tendency across all subjects and their GPA (Grade Point Average) as feature quantities, and calculates a correction value corresponding to the cluster to which Learner A belongs. In this embodiment, the process here simply determines which cluster Learner A belongs to, and the clustering itself is performed in advance using a method such as k-means. Note that this cluster classification may be given or set in advance, or the individual learning tendency analysis unit 27 may have a clustering function during learning. In the example shown in FIG. 8A , the individual learning tendency analysis unit 27 assigns four clusters: "Read lecture materials carefully and get good grades: correction value 1," "Do not read lecture materials and get bad grades: correction value 0.7," "Read lecture materials carefully but get bad grades: correction value 0.5," and "Do not read lecture materials and get good grades: correction value 0.3." The cluster to which learner A belongs belongs to the "reads lecture materials carefully and gets good grades" category out of the four categories, and the correction value is "1".
[0070] 4) Calculate the product of the median viewing time of all learners calculated in 2) above / the viewing time of learner A and the correction value derived in 3) to obtain "P read (lecture materials, page, learner A's ID)
[0071] 5) The "I" calculated in 1) above read (Lecture materials, pages)" and the "P" calculated in 4) above. read (lecture material, page, learner A's ID)) and read (Lecture materials, page) P read (Lecture materials, page)
[0072] Next, in the third term on the right side of the above equation (3), "I test (Lecture materials, page) P test As shown in FIG. 9, the following calculations are performed for (lecture material, page, learner A's ID) using steps 6) to 9).
[0073] 6) Determine the relevant questions for each page of the lecture materials. If multiple questions are related to one page, select the question with the highest overall average score, or the question that Learner A got wrong, etc., to determine one question per page.
[0074] 7) The overall learning tendency analysis unit 25 calculates the average score for each question for all learners and outputs the result as "I test (Lecture materials, page)
[0075] 8) The individual learning tendency analysis unit 27 calculates the perfect score / score of learner A for each question and outputs "P test (lecture materials, page, learner A's ID)
[0076] 9) The "I" calculated in 7) above test (Lecture materials, pages)" and the "P" calculated in 8) above. test (lecture material, page, learner A's ID)) and test (Lecture materials, page) P test (lecture materials, page, learner A's ID)
[0077] Next, using the above formula (3), the importance calculation unit 36 calculates the importance I (lecture material, page, learner A's ID) for each page of the lecture material.
[0078] The importance calculation unit 36 refers to the weight information 56B stored in the storage 54 and specifies the weight of each feature. An example of the weight information 56B is shown in FIG. 10A. The weight information 56B shown in FIG. 10A represents the importance of each feature in a grade prediction model 56A that predicts the final grade of a target course, characterized by each learning behavior and each learning result. In the example shown in FIG. 10A, the weight ω read is set to "0.30", and the weight ω test is set to "0.60".
[0079] Furthermore, the importance calculation unit 36 uses the weights of the features identified above and the above equation (3) to calculate the importance I (lecture material, page, learner A's ID) for each page of the lecture material, as shown in Figure 10B.
[0080] Therefore, for each feature amount k, the importance calculation unit 36 calculates a multiplication value by multiplying the importance based on the overall learning tendency related to the feature amount k, the importance based on the individual learning tendency related to the feature amount k, and the weight of the feature amount. read (Lecture materials, page) P read (lecture material, page, learner A's ID)) with weight "ω read ". The importance calculation unit 36 also multiplies "I test (Lecture materials, page) P test (lecture material, page, learner A's ID)) with weight "ω test " and the importance of the lecture material "I content The importance of each page is calculated by adding "(lecture material, page)" and the multiplied value.
[0081] For example, if the lecture material is "teaching material A," the id of learner A is "0005," and the calculation result shown in Figure 10B is obtained, the importance of page 13 of teaching material A (lecture material) is derived as shown in equation (4) below.
[0082] I (material A, 12,0005) = 1.00 + 0.30 x 0.50 + 0.60 x 1.0 = 1.75 ... (4)
[0083] That is, the importance of page 12 of learning material A for learner A is derived to be "1.75."
[0084] The teaching material providing unit 38 extracts pages to be provided in descending order of the importance of each page for each learner, as calculated by the importance calculating unit 36. For example, if it is predetermined that the seven most important pages of material should be provided as personalized teaching materials, the teaching material providing unit 38 extracts seven pages of material from the lecture materials in descending order of importance.
[0085] Then, each extracted page is arranged based on a predetermined format and provided to the target learner. The predetermined format may be, for example, a format in which thumbnail images of all extracted pages are generated and displayed as a list, but is not particularly limited thereto and may be any format, and a format appropriate for the learner may be applied. The method of providing the information is also not particularly limited and may be a format in which a printout is sent to the learner, a format in which the information is displayed on an electronic dashboard used by each learner, or a format in which the generated digital data is sent to the learner by email or the like, and is not particularly limited thereto. The method of providing the information may be in a format desired by the learner.
[0086] As described above, in the above-described learning material providing device 20, the learning material analysis unit 23 analyzes the learning material 10 and derives a learning material-specific importance for each range of the learning material 10. The overall learning tendency analysis unit analyzes the learning tendencies of multiple learners who are using the learning material 10 and derives an importance for each range of the learning material 10 based on the overall learning tendency. The individual learning tendency analysis unit 27 analyzes the learning tendencies of individual learners to whom the learning material 10 is to be provided and derives an importance for each range of the learning material 10 based on the individual learning tendency. The importance calculation unit 36 calculates an importance for each range of the learning material 10 using the learning material-specific importance, the importance based on the overall learning tendency, the importance based on the individual learning tendency, and the feature weight ω obtained by the grade prediction model 56A that uses learning behavior or learning results as features. The learning material providing unit 38 then provides the range of the learning material 10 extracted from the learning material 10 based on the importance to the learner to whom the learning material 10 is to be provided.
[0087] Therefore, the above-described teaching material providing device 20 can provide personalized teaching materials to target students.
[0088] In the above embodiment, the weight information 56B is derived by the learning material providing device 20, i.e., the above-described preparation process is performed by the learning material providing device 20. However, the preparation process may be performed by a device separate from the learning material providing device 20. That is, the above-described weight determination process (see FIG. 4) and the text provision process (see FIG. 6) may be performed by separate devices.
[0089] Furthermore, the various processes executed by the CPU after reading the software (program) in each of the above embodiments may be executed by various processors other than the CPU. Examples of processors in this case include dedicated electrical circuits, such as programmable logic devices (PLDs) whose circuit configuration can be changed after manufacture, such as field-programmable gate arrays (FPGAs), and application-specific integrated circuits (ASICs), which are processors having a circuit configuration designed specifically for executing specific processes. Each of the learning process and the estimation process may be executed by one of these various processors, or by a combination of two or more processors of the same or different types (e.g., multiple FPGAs, a combination of a CPU and an FPGA, etc.). The hardware structure of these various processors is, more specifically, an electrical circuit that combines circuit elements such as semiconductor elements.
[0090] In addition, in the above-described embodiments, the weight determination program 55A and the learning material provision program 55B are each pre-stored (installed) in the storage 54, but this is not limiting. The weight determination program 55A and the learning material provision program 55B may each be provided in a form stored on a non-transitory storage medium such as a CD-ROM (Compact Disk Read Only Memory), a DVD-ROM (Digital Versatile Disk Read Only Memory), or a USB (Universal Serial Bus) memory. Furthermore, the weight determination program 55A and the learning material provision program 55B may each be downloaded from an external device via a network.
[0091] The following additional notes are provided regarding the above-described embodiments.
[0092] (Supplementary Note 1) A teaching material providing device comprising: a teaching material analysis unit that analyzes teaching materials and derives a teaching material-specific importance for each range of the teaching materials; an overall learning tendency analysis unit that analyzes the learning tendencies of multiple learners who are using the teaching materials together and derives a level of importance based on the overall learning tendency for each range of the teaching materials; an individual learning tendency analysis unit that analyzes the learning tendencies of individual learners to whom the teaching materials are to be provided and derives a level of importance based on the individual learning tendency for each range of the teaching materials; an importance calculation unit that calculates an importance for each range of the teaching materials using the teaching material-specific importance, the importance based on the overall learning tendency, the importance based on the individual learning tendency, and weights of the features obtained by a grade prediction model that uses learning behavior or learning results as features; and a teaching material providing unit that provides the range of teaching materials extracted from the teaching materials based on the importance to the learners to whom the teaching materials are to be provided.
[0093] (Supplementary Note 2) The teaching material providing device described in Supplementary Note 1, wherein the overall learning trend analysis unit derives, for each of the feature amounts, an importance based on the overall learning trend associated with the feature amount; the individual learning trend analysis unit derives, for each of the feature amounts, an importance based on the individual learning trend associated with the feature amount; and the importance calculation unit multiplies, for each of the feature amounts, the importance based on the overall learning trend associated with the feature amount, the importance based on the individual learning trend associated with the feature amount, and the weight of the feature amount to calculate a multiplied value, and adds the importance specific to the teaching material and the multiplied value calculated for each of the feature amounts to calculate the importance of the range.
[0094] (Supplementary Note 3) The teaching material providing device according to Supplementary Note 1 or Supplementary Note 2, wherein the weight is a contribution of the feature obtained by the grade prediction model.
[0095] (Supplementary Note 4) A method of providing teaching materials, comprising: analyzing teaching materials and deriving a teaching material-specific importance for each range of the teaching materials; analyzing the learning tendencies of multiple learners who are using the teaching materials as a whole and deriving a importance based on the overall learning tendency for each range of the teaching materials; analyzing the learning tendencies of individual learners to whom the teaching materials are to be provided and deriving a importance based on the individual learning tendency for each range of the teaching materials; calculating an importance for each range of the teaching materials using the teaching material-specific importance, the importance based on the overall learning tendency, the importance based on the individual learning tendency, and weights of the features obtained from a grade prediction model that uses learning behavior or learning results as features; and providing the range of teaching materials extracted from the teaching materials based on the importance to the learners to whom the teaching materials are to be provided.
[0096] (Supplementary Note 5) A program for causing a computer to function as each part of the educational material providing device according to any one of Supplementary Note 1 to Supplementary Note 3.
[0097] (Supplementary Note 6) A non-transitory storage medium storing a program executable by a computer to execute a teaching material provision process, wherein the teaching material provision process: analyzes teaching materials and derives a teaching material-specific importance for each range of the teaching materials; analyzes the learning tendencies of multiple learners who are using the teaching materials as a whole and derives a importance based on the overall learning tendency for each range of the teaching materials; analyzes the learning tendencies of individual learners to whom the teaching materials are to be provided and derives a importance based on the individual learning tendency for each range of the teaching materials; calculates an importance for each range of the teaching materials using the importance based on the overall learning tendency, the importance based on the individual learning tendency, and weights of the features obtained from a grade prediction model in which learning behavior or learning results are used as features; and provides the range of teaching materials extracted from the teaching materials based on the importance to the learners to whom the teaching materials are to be provided.
[0098] (Supplementary Note 7) A teaching material providing device comprising: a memory; and at least one processor connected to the memory, wherein the processor is configured to: analyze teaching materials and derive a teaching material-specific importance for each range of the teaching materials; analyze the learning tendencies of a plurality of learners who are using the teaching materials as a whole and derive a importance based on the overall learning tendency for each range of the teaching materials; analyze the learning tendencies of individual learners to whom the teaching materials are to be provided and derive a importance based on the individual learning tendency for each range of the teaching materials; calculate an importance for each range of the teaching materials using the importance based on the overall learning tendency, the importance based on the individual learning tendency, and weights of the features obtained by a grade prediction model using learning behavior or learning results as features; and provide the range of teaching materials extracted from the teaching materials based on the importance to the learners to whom the teaching materials are to be provided.
[0099] 10 Teaching material, 11 Personalized teaching material 20 Teaching material providing device 23 Teaching material analysis unit 25 Overall learning tendency analysis unit 27 Individual learning tendency analysis unit 30 Score prediction model construction unit 32 Weight determination unit 51 CPU 55B Teaching material providing program 56A Score prediction model, 56B Weight information
Claims
1. A teaching material provision device comprising: a teaching material analysis unit that analyzes teaching materials and derives a teaching material-specific importance for each range of the teaching material; an overall learning tendency analysis unit that analyzes the learning tendencies of a plurality of learners who are using the teaching material and derives an importance based on the overall learning tendency for each range of the teaching material; an individual learning tendency analysis unit that analyzes the learning tendencies of each individual learner to whom the teaching material is to be provided and derives an importance based on the individual learning tendency for each range of the teaching material; an importance calculation unit that calculates an importance for each range of the teaching material using the teaching material-specific importance, the importance based on the overall learning tendency, the importance based on the individual learning tendency, and weights of the features obtained by a grade prediction model in which learning behavior or learning results are used as features; and a teaching material provision unit that provides the range of teaching material extracted from the teaching material based on the importance to the learner to whom the teaching material is to be provided.
2. The educational material providing device of claim 1, wherein the overall learning trend analysis unit derives, for each of the features, an importance based on the overall learning trend associated with the feature, the individual learning trend analysis unit derives, for each of the features, an importance based on the individual learning trend associated with the feature, and the importance calculation unit multiplies, for each of the features, the importance based on the overall learning trend associated with the feature, the importance based on the individual learning trend associated with the feature, and the weight of the feature to calculate a multiplied value, and adds the importance specific to the educational material and the multiplied value calculated for each of the features to calculate the importance of the range.
3. The educational material providing device according to claim 1, wherein the weight is a contribution of the feature obtained by the grade prediction model.
4. A method for providing teaching materials, comprising: analyzing teaching materials and deriving a teaching material-specific importance for each range of the teaching materials; analyzing the learning tendencies of a plurality of learners using the teaching materials as a whole and deriving a importance based on the overall learning tendency for each range of the teaching materials; analyzing the learning tendencies of each individual learner to whom the teaching materials are to be provided and deriving a importance based on the individual learning tendency for each range of the teaching materials; calculating an importance for each range of the teaching materials using the teaching material-specific importance, the importance based on the overall learning tendency, the importance based on the individual learning tendency, and weights of the features obtained by a grade prediction model using learning behavior or learning results as features; and providing the range of the teaching materials extracted from the teaching materials based on the importance to the learner to whom the teaching materials are to be provided.
5. A program for causing a computer to function as each part of the teaching material providing device according to any one of claims 1 to 3.
Citation Information
Patent Citations
Learning system
JP2002278439A
Learning support system and learning support method, and learning support program and recording medium
JP2003098948A
Learner acquisition characteristic analysis system, its method, and program
JP2006072122A
How to augment lectures based on previous exams
JP2008535017A
E-learning system
JP2019124964A
