JP7797846B2 - Information processing device and program - Google Patents

Description

The present invention relates to technology for outputting information related to learning.

Systems that evaluate learners' answers on a computer have been known for some time. Patent Document 1 discloses a learning device that stores the content entered by the learner in an answer area and performs a correct/incorrect evaluation to detect which parts of the calculation process contain errors.

JP 2018-87982 A

However, the invention described in Patent Document 1 merely evaluates the correctness of answers based on the content of the answers provided by the learner and detects parts that contain errors, and is not intended to evaluate whether the content of the answers is appropriate or provide information accordingly.

The present invention has been made to solve the above-mentioned problems, for example, by evaluating the content of learners' answers and providing appropriate information to learners and learning support staff.

In one aspect of the present invention, an information processing device is communicatively connected to a terminal device used by a target learner who is a specific learner among a plurality of learners, and comprises: a learning history memory unit that stores answer information regarding answers to questions by a plurality of learners including the target learner; a target question acquisition unit that acquires information regarding a target question, which is a question specified by the target learner, from the terminal device; an extraction unit that extracts answer information to the target question by a plurality of learners including the target learner from the learning history memory unit; a group classification unit that classifies the answers to the target question by each of the plurality of learners including the target learner into groups subdivided by answer content patterns based on the answer information extracted by the extraction unit; and a group identification unit that identifies a group to which a target answer, which is the answer to the target question by the target learner, belongs based on target answer information, which is the answer information to the target question by the target learner.

In one aspect of the information processing device described above, group information regarding the group to which the target answer belongs is output. According to this aspect, the information processing device outputs information regarding the group to which the subject's answer to the target question belongs, and can provide this information to the subject or a learning supporter.

One aspect of the above information processing device includes a correct answer rate calculation unit that calculates the correct answer rate of answers that belong to the same group as the target answer based on the answer information extracted by the extraction unit, and the group information output unit outputs the correct answer rate as the group information. According to this aspect, the information processing device outputs the correct answer rate of the group to which the subject's answer to the target question belongs, and provides this to the subject and a learning supporter. In other words, the information processing device can predict the correct answer rate of the subject's answer to the target question from the pattern of the answer content, and therefore can evaluate whether the answer content is appropriate.

One aspect of the information processing device described above includes an academic ability calculation unit that calculates an academic ability evaluation value of learners who have worked on answers that belong to the same group as the target answer, based on the answer information extracted by the extraction unit, and the group information output unit outputs the academic ability evaluation value as the group information. According to this aspect, the information processing device outputs the academic ability evaluation value of the group to which the subject's answer to the target question belongs, and provides it to the subject or a learning supporter. This allows the subject and a learning supporter to confirm the subject's academic ability level according to the content of the subject's answer.

In one aspect of the information processing device described above, other group information regarding other groups to which the target answer does not belong is output. According to this aspect, the information processing device can output information regarding other groups and provide it to the target person or a learning supporter.

One aspect of the above information processing device includes a correct answer rate calculation unit that calculates the correct answer rate of answers belonging to the other group based on the answer information extracted by the extraction unit, and the other group information output unit outputs the correct answer rate as the other group information. According to this aspect, the subject and/or learning supporter can compare the correct answer rate of the group to which the subject's answer to the target question belongs with the correct answer rates of the other groups.

One aspect of the information processing device described above includes an academic ability calculation unit that calculates an evaluation value of the academic ability of the learner who worked on the answer belonging to the other group based on the answer information extracted by the extraction unit, and the other group information output unit outputs the evaluation value of the academic ability as other group information. According to this aspect, the subject and/or learning supporter can compare the evaluation value of the academic ability of the group to which the subject's answer to the target question belongs with the evaluation value of the academic ability of the other groups.

In one aspect of the information processing device described above, the other group information output unit outputs answers belonging to the other group as the other group information based on the answer information extracted by the extraction unit. According to this aspect, the subject and/or learning supporter can study by comparing the subject's answers with the answers of the other groups.

In another aspect of the information processing device, the questions are linked to information on learning elements that indicate learning content, and a similar problem identification unit is provided that identifies questions with the same learning elements as similar questions, the extraction unit further extracts from the learning history storage unit learners who have tackled the similar questions and answer information related to answers to the similar questions, and the group classification unit classifies answers to the target question and the similar questions by each of multiple learners including the target person into groups subdivided by patterns of answer content based on the extracted answer information. According to this aspect, when only a few learners have tackled the target question, the information processing device can identify similar questions and classify the answers to the target question and the similar questions by each learner into groups subdivided by patterns of answer content.

In another aspect of the present invention, there is provided a program executed by an information processing device having a computer, the information processing device being communicatively connected to a terminal device used by a target learner who is a specific learner among a plurality of learners, and causing the computer to function as: a learning history storage unit that stores answer information regarding answers to questions by a plurality of learners including the target learner; a target question acquisition unit that acquires, from the terminal device, information regarding a target question that is a question specified by the target learner; an extraction unit that extracts answer information to the target question by a plurality of learners including the target learner from the learning history storage unit; a group classification unit that classifies the answers to the target question by each of the plurality of learners including the target learner into groups subdivided by answer content patterns based on the answer information extracted by the extraction unit; and a group identification unit that identifies a group to which a target answer that is the answer to the target question by the target learner belongs based on target answer information that is the answer information to the target question by the target learner. By installing and executing this program on a computer, the information processing device of the present invention can be configured.

The information processing device according to the present invention can evaluate the content of a learner's answers and provide appropriate information to the learner and their learning support staff.

1 shows the configuration of a learning support system to which a server of the present invention is applied. FIG. 2 is a block diagram showing the hardware configuration of a server. 10 is an example of the data configuration of a learning element DB. 10 is an example of a data structure of a question DB. 10 is an example of a data configuration of a learning history DB. FIG. 2 is a block diagram showing the functional configuration of a server according to the first embodiment. These are sample answers that each student wrote in the answer section of the target question. 10A and 10B are diagrams illustrating a method for extracting sentences and formulas from answer information. FIG. 10 is a diagram illustrating the criteria for determining one continuous calculation formula. FIG. 10 is a diagram illustrating a method for extracting feature quantities of an answer based on a text. FIG. 10 is a diagram illustrating a method for grouping answers to target questions by each student. 1 is an example of a feature space in which answers to target questions by subjects are plotted. 10A and 10B are examples of a prediction screen and a recommendation screen according to the first embodiment. FIG. 2 is a block diagram showing the hardware configuration of a student terminal. 10 is a flowchart of a screen creation process according to the first embodiment. FIG. 10 is a block diagram showing the functional configuration of a server according to a second embodiment. 10 is an example of an LRT evaluation value for each group. 10A and 10B are examples of a prediction screen and a recommendation screen according to the second embodiment. 10 is a flowchart of a screen creation process according to a second embodiment. 13 shows an example of the configuration of a learning support system according to a fourth modified example.

Hereinafter, an embodiment of the present invention will be described with reference to the drawings.
First Embodiment
[Overall configuration]
1 shows the configuration of a learning support system to which the server of the present invention is applied. The learning support system 100 is a system that predicts the correct answer rate based on the content of a learner's answers and recommends appropriate answer examples. The learning support system 100 is configured so that a server 10, multiple student terminals 20, and multiple teacher terminals 30 can communicate with each other via a network 5 such as the Internet.

The server 10 is an information processing device that processes, stores, and transmits and receives various types of information, and is, for example, a server device, a personal computer, or a general-purpose tablet PC (personal computer).

Student terminals 20 are used by learners such as students and their parents, and are information processing devices such as smartphones, mobile phones, wearable devices such as Apple Watch (registered trademark), tablets, and personal computer terminals. Specifically, student terminals 20 transmit answer information and target question information, and receive and display prediction screens and recommendation screens.

The teacher terminal 30 is used by a learning supporter such as a teacher, and is, for example, an information processing device such as a tablet or personal computer terminal. Specifically, the teacher terminal 30 transmits target question information, receives and displays prediction screens and recommendation screens, etc.

[Server configuration]
2 is a block diagram showing the hardware configuration of server 10. Server 10 includes a communication unit 11, a control unit 12, a storage unit 13, a recording medium 14, a display unit 15, and an input unit 16. These components, a learning element database (hereinafter, "database" will be referred to as "DB") 41, a question DB 42, and a learning history DB 43 are interconnected via a bus 19.

Note that the server 10 may be run on a single computer, distributed across multiple computers, or distributed across virtual machines.

The communication unit 11 is a communication unit for communicating with the student terminal 20 and the teacher terminal 30 via the network 5. Specifically, the communication unit 11 receives target question information from the student terminal 20 and the teacher terminal 30, and sends prediction screens and recommendation screens to the student terminal 20 and the teacher terminal 30.

The control unit 12 includes processing units such as a CPU (Central Processing Unit), MPU (Micro-Processing Unit), and GPU (Graphics Processing Unit), and performs various information processing, control processing, and other operations related to the server 10 by reading and executing programs stored in the memory unit 13. The programs can be deployed to run on a single computer, at a single site, or distributed across multiple sites and on multiple computers interconnected by a communications network. While Figure 2 illustrates the control unit 12 as a single processor, it may also be a multi-processor.

The memory unit 13 includes memory elements such as RAM (Random Access Memory) and ROM (Read Only Memory), and stores programs, data, etc. necessary for the control unit 12 to execute processing. The memory unit 13 also temporarily stores data, etc. necessary for the control unit 12 to execute calculations.

The recording medium 14 is a non-volatile, non-transitory recording medium such as a disk-shaped recording medium or semiconductor memory, and is configured to be detachable from the server 10. The recording medium 14 stores various programs executed by the control unit 12. When the server 10 executes the screen creation process described below, the programs stored on the recording medium 14 are loaded into the memory unit 13 and executed by the control unit 12.

The display unit 15 is a liquid crystal display or organic EL (electroluminescence) display, etc., and displays various information according to instructions from the control unit 12. The input unit 16 is an input device such as a mouse, keyboard, touch panel, or buttons, and outputs received operation information to the control unit 12.

The learning element DB41 stores information about learning elements, which are subdivisions of learning content. Here, learning elements are learning content and textbook explanations that have been subdivided to an optimal level of granularity, taking into account learning groups such as units, for purposes such as linking textbooks and teaching materials, understanding learning progress, and creating lesson plans. Furthermore, a unit is a group of learning activities centered around a specific theme, and is set for each subject, such as "Mathematics" or "Japanese," and is composed of one or more learning elements.

Figure 3 is an example of the data structure of the learning element DB41. As shown, the learning element DB41 stores information on subjects, fields, units, learning elements, and learning element IDs. The subject is the name of the subject, such as "Mathematics" or "Japanese." The field is the name of the field, set for each subject, such as "Differentiation." The unit is the name of the unit, set for each field, such as "Differentiation of Composite Functions." The learning element is information indicating the learning content that the learning element targets, set for each unit, such as "Logarithm" or "Negative Order." The learning element ID is identification information for the learning element.

It is assumed that the fields that make up a subject, the units that make up a field, and the learning elements that make up a unit are each set in advance. The data structure of the learning element DB41 shown in Figure 3 is an example, and the present invention is not limited to this. The data structure of the learning element DB41 can be set arbitrarily as long as it stores information related to the identification of learning elements.

Question DB 42 stores question information about questions that students work on, in association with question IDs that identify the questions. Figure 4 shows an example of the data structure of question DB 42. As shown in the figure, question DB 42 stores information such as question IDs, question data, learning element IDs, learning material names, pages, and difficulty levels. Question data is data about the question indicated by the question ID. Learning element ID is identification information for the learning element to which the question indicated by the question ID belongs. Learning material name is the name of the learning material, such as a reference book or textbook, that contains the problem indicated by the question ID, and page is the page and question number on which the problem appears in the learning material.

The difficulty level is a numerical value indicating the difficulty of the question indicated by the question ID. Here, the question difficulty level (hereinafter simply referred to as "difficulty level") is based on any concept, such as the difficulty level of the question or the correct answer rate, and may be set by evaluating each question using Item Response Theory (IRT) or the like, or may be set artificially in advance. Specifically, in this embodiment, the difficulty level is calculated numerically based on the student's answer information, which will be described later. The lower the numerical value, the easier the difficulty level, and the higher the numerical value, the more difficult the difficulty level. Note that the difficulty level may be stored in the question DB 42 as shown in Figure 4 and extracted when needed, or it may not be stored in the question DB 42 and instead be calculated based on the answer information each time it is needed.

Note that which learning element a question belongs to is set in advance. Furthermore, the data structure of question DB42 shown in Figure 4 is an example, and the present invention is not limited to this, and can be set arbitrarily.

The learning history DB43 stores answer information related to the answers to questions that students have worked on, in association with a student ID that identifies the student. Figure 5 shows an example of the data structure of the learning history DB43. As shown in the figure, the learning history DB43 stores student IDs, question IDs, answer column data, and correct/incorrect information. The question ID is identification information for the question that the student identified by the student ID worked on. The answer column data is data that the student entered in the answer column as their answer to the question. Correct/incorrect information is whether the answer to the question that the student identified by the student ID worked on was correct or incorrect.

In this embodiment, as an example, the student terminal 20 is linked in advance to the student ID of the student who uses it, and answer information regarding the answers to questions that the student has worked on using the student terminal 20 is automatically stored in the learning history DB 43 of the server 10. Furthermore, the correctness of the answers included in the answer information may be determined, for example, by the server 10 storing model answers for each question in advance and automatically scoring based on the answer column data, or by obtaining the teacher's scoring results. Furthermore, the data structure of the learning history DB 43 shown in Figure 5 is an example, and the present invention is not limited to this, and the data structure can be set as desired.

Note that the storage format of each DB described above is an example, and other storage formats are possible as long as the relationships between data are maintained. Each DB is implemented using a recording medium such as an HDD (Hard Disk Drive) or SSD (Solid State Drive). In this embodiment, the storage unit 13 and the various DBs may be configured as an integrated storage device, or may be separate storage devices. The various DBs may also be external storage devices connected to the server 10, and the configuration can be set as desired.

Figure 6 is a block diagram showing the functional configuration of the server. Functionally, the server 10 includes a learning element DB 41, a question DB 42, a learning history DB 43, a target question acquisition unit 51, an extraction unit 52, a group classification unit 53, a correct answer rate calculation unit 54, a group identification unit 55, and a screen creation unit 56.

The target question acquisition unit 51, extraction unit 52, group classification unit 53, correct answer rate calculation unit 54, group identification unit 55, and screen creation unit 56 are realized by the control unit 12 executing a program.

The target question acquisition unit 51 acquires target question information including a student ID and information related to the target question (hereinafter also referred to as the "target question") from the student terminal 20. When a student wants to check the accuracy rate of the answer they entered in the answer field or refer to other example answers, they use the student terminal 20 to specify a specific question as the target question and send the target question information to the server 10. The target question acquisition unit 51 acquires the target question information from the student terminal 20 and designates the student as the target.

The extraction unit 52 extracts from the learning history DB 43 the students who worked on the target questions and answer information related to the answers given by each student to the target questions. Note that the students who worked on the target questions extracted by the extraction unit 52 include the target person.

The group classification unit 53 classifies each student's answers to the target questions into groups subdivided according to the patterns of answer content, based on the answer information extracted by the extraction unit 52.

Specifically, the classification method used by the group classification unit 53 will be described. Figure 7 shows example answers written by each student in the answer column of the target problem. Figure 7(a) shows an example answer written by student A in answer column 61 of target problem 60. The pattern of student A's answer, as shown, is "includes intermediate steps, with supplementary explanation in Japanese." Figure 7(b) shows an example answer written by student B in answer column 62 of target problem 60. The pattern of student B's answer, as shown, is "includes intermediate steps, no supplementary explanation in Japanese." Figure 7(c) shows an example answer written by student C in answer column 63 of target problem 60. The pattern of student C's answer, as shown, is "no intermediate steps, no supplementary explanation in Japanese."

Figure 8 is a diagram explaining a method for extracting text and formulas from answer information. Figure 9 is a diagram explaining the criteria for determining a single continuous formula (hereinafter also referred to as "one formula"). The group classification unit 53 first extracts text other than formulas from the answer column data included in the answer information. Specifically, the group classification unit 53 extracts text 71 other than formulas and text 72, as shown in Figure 8(a). Next, the group classification unit 53 extracts formulas from the answer column data. Specifically, the group classification unit 53 detects "=" (equal) and treats all terms connected by "=" as one formula, extracting formulas 73 to 77 as shown in Figure 8(b). At this time, the group classification unit 53 treats terms with a certain number of spaces or more as separate formulas. For example, because there are more than a certain number of blank spaces between formulas 73 and 74, the group classification unit 53 determines that they are separate formulas rather than a single formula, and extracts formulas 73 and 74 from the answer column data. Furthermore, as shown in FIG. 9, in the case of horizontally consecutive formulas that do not contain more than a certain number of blank spaces, the group classification unit 53 determines that all terms connected by "=" are a single consecutive formula, and extracts, for example, formula 77 from the answer column data. Furthermore, as shown in FIG. 9, in the case of vertically consecutive formulas 75a and 75b, if there is no term to the left of the "=" in formula 75b, the group classification unit 53 determines that it is a single formula connected to the above formula 75a, and extracts, for example, formula 75 from the answer column data.

Figure 10 is a diagram illustrating a method for extracting answer features based on text. The group classification unit 53 first combines the extracted sentences and formulas into a single text. Specifically, as shown in Figure 10(a), the group classification unit 53 combines the extracted sentences 71 and 72 with formulas 73 to 77 into a single text 78.

The group classification unit 53 then extracts text features using a machine learning technique capable of processing time-series information and text, such as a Long Short Term Memory (LSTM) as shown in FIG. 10(b). Note that if the LSTM's input and output are the same text and the learning is performed to have "input," "feature extraction in the intermediate layer," and "output that restores the input," the intermediate layer features can be acquired in the process. Therefore, for example, the group classification unit 53 can acquire the feature values of text 78 in the intermediate layer by learning to restore and output the exact same text 78 using text 78 as input. Here, the feature values of text 78 are the answer column data corresponding to text 78, i.e., the feature values of each student's answer to the target problem. In other words, the group classification unit 53 acquires the feature values of each student's answer to the target problem based on the answer information of the target problem that each student has worked on. Specifically, the feature values represent patterns of answer content, such as whether or not there are intermediate steps and whether or not there are supplementary explanations in Japanese.

Figure 11 is a diagram illustrating a method for grouping answers to target questions by each student. As shown in Figure 11(a), the group classification unit 53 plots each student's answer to the target question in a feature space based on the feature. In other words, the plotted circles represent the student's answer to the target question. The group classification unit 53 then groups groups of answers that tend to have similar answer content patterns. Examples of grouping methods include a combination of RANSAC (Random Sample Consensus) and k-means. Specifically, as shown in Figure 11(b), the group classification unit 53 groups the circles representing each student's answer plotted in the feature space into Group A, Group B, Group C, and Group D by clustering. In this way, the group classification unit 53 classifies each student's answer to the target question into groups subdivided by answer content pattern based on the answer information extracted by the extraction unit 52.

As shown in Figure 11(c), the correct answer rate calculation unit 54 calculates the correct answer rate for each group based on the answer information extracted by the extraction unit 52.

The group identification unit 55 identifies the group to which the subject's answer to the target question (hereinafter also referred to as the "target answer") belongs, based on answer information regarding the subject's answer to the target question (hereinafter also referred to as the "target answer information"). Figure 12 is an example of a feature space in which the target answer is plotted. Specifically, first, the group classification unit 53 acquires the feature quantities of the target answer as numerical values based on the target answer information, and plots the target answer with a star mark in the feature space, as shown in Figure 12. The group identification unit 55 identifies the group in the feature space with the closest distance between the star mark and the center of gravity of each group as the group to which the target answer belongs. In the case of a feature space such as that shown in Figure 12, the group identification unit 55 identifies the group to which the target answer belongs as Group A.

When the screen creation unit 56 acquires the target question information, it references each DB and creates and outputs a prediction screen that allows the user to check the predicted accuracy rate (hereinafter also referred to as the "predicted accuracy rate") for the answer they entered in the answer field.

Here, we will explain the prediction screen created by the screen creation unit 56. Figure 13(a) is an example of a prediction screen. The prediction screen is a screen that displays the predicted correct answer rate of the target answer, and is transmitted to the student terminal 20 of the subject. The subject can check the correct answer rate of their own prediction of the target answer by displaying the prediction screen received from the server 10 using the student terminal 20. As shown in Figure 13(a), the prediction screen has a subject item 79, a learning element item 80, a question specification item 81, a target question 82, an answer column 83, a group correct answer rate item 84, and another group's correct answer rate item 85.

The target student item 79 displays the student ID of the target student. The learning element item 80 displays the unit and learning element corresponding to the target question. The question specification item 81 has a question item 81a that displays the target question specified by the target student and its difficulty level, and question items 81b and 81c that display questions of the same learning element as the target question and their difficulty level. A check mark is displayed on the question item 81a that displays the target question.

The target student may use the student terminal 20 to perform a predetermined operation to display a prediction screen that includes only the target student item 79, learning element item 80, and question specification item 81. In this case, the target student can use the student terminal 20 to specify the target question by selecting a predetermined learning element and question in the learning element item 80 and question specification item 81, respectively, and transmit target question information including the student ID and information related to the target question to the server 10.

The target question 82 displays the target question specified by the subject. The answer column 83 displays the target answer. The group correct answer rate item 84 displays the correct answer rate of the group to which the target answer belongs. Because the group to which the target answer belongs has similar answer content patterns, the correct answer rate of the target answer is predicted to be similar to the correct answer rate of the group. In other words, the predicted correct answer rate of the target answer is the correct answer rate of the group to which the target answer belongs. The other group correct answer rate item 85 displays the correct answer rates of other groups to which the target answer does not belong, and has a link that can be clicked to display answers belonging to other groups.

Specifically, if students' answers to a target question are classified into four groups, A to D, and the target answer belongs to group C, the group correct answer rate item 84 will display group C's correct answer rate of 85% as "Correct answer rate for same answer pattern: 85%," as shown in FIG. 13(a). Similarly, the other group correct answer rate item 85 will display group A's correct answer rate of 93%, as shown in FIG. 13(a), along with a link to display answers belonging to group A, as "View example answers with a 93% correct answer rate." Similarly, the other group correct answer rate item 85 will display the correct answer rates for group B and group D, as "View example answers with a 60% correct answer rate" and "View example answers with a 30% correct answer rate," respectively, along with links to display answers belonging to each group.

When the screen creation unit 56 receives information from the student terminal 20 that a specific link has been clicked by the target student, it creates and outputs a recommendation screen including the answers of students belonging to the group corresponding to the link.

Here, we will explain the recommendation screen created by the screen creation unit 56. Figure 13(b) is an example of a recommendation screen. The recommendation screen is a screen that displays answers belonging to other groups, and is sent to the student terminal 20 of the target student. By displaying the recommendation screen received from the server 10 using the student terminal 20, the target student can refer to answers from other groups whose answer content patterns are different from their own. As shown in Figure 13(b), the recommendation screen has a target student item 79, a learning element item 80, a question specification item 81, a target question 82, an answer column 86, and a group correct answer rate item 87.

The target person item 79, learning element item 80, question specification item 81, and target question 82 are the same as those on the prediction screen, and therefore will not be described for the sake of brevity. The answer column 86 displays answers belonging to other groups to which the target answer does not belong. Specifically, the screen creation unit 56 identifies the group corresponding to the link clicked by the target in the other group correct answer rate item 85 on the prediction screen and creates a recommendation screen to display answers belonging to that group. For example, if the target uses the student device 20 to click "View example answers with a 93% correct answer rate" in the other group correct answer rate item 85 on the prediction screen, as shown in FIG. 13(a), the screen creation unit 56 identifies group A, which has a correct answer rate of 93%, and creates a recommendation screen to display answers belonging to group A in the answer column 86. The answer column 86 can display multiple answers belonging to group A using scroll buttons, next buttons, etc. The display order of the multiple answers belonging to Group A can be set arbitrarily, such as alphabetical order by student name or by student academic level. The group correct answer rate item 87 displays the correct answer rate of the group to which the answer displayed in the answer column 86 belongs.

In this embodiment, the other group correct answer rate item 85 on the prediction screen displays the correct answer rates and links for all other groups to which the target answer does not belong. However, the present invention is not limited to this, and it may also be possible to display the correct answer rates and links for only other groups with a higher correct answer rate than the correct answer rate of the group to which the target answer belongs, or for only the other group with the highest correct answer rate. Specifically, the other group correct answer rate item 85 may display the correct answer rate and link for the other group with the highest correct answer rate, along with the sentence recommended as study support.

In addition, the screen creation unit 56 may not provide the other group correct answer rate item 85 on the prediction screen, but instead display the correct answer rate and a link in the group correct answer rate item 84, and when the link is clicked, create a recommendation screen that allows reference to other answers that belong to the same group as the target answer.

In the above configuration, the learning history DB 43 of the server 10 is an example of the learning history storage unit of the present invention. Also, in the above configuration, the target question acquisition unit 51, extraction unit 52, group classification unit 53, correct answer rate calculation unit 54, and group identification unit 55 of the server 10 are examples of the target question acquisition unit, extraction unit, group classification unit, correct answer rate calculation unit, and group identification unit of the present invention, respectively. Also, in the above configuration, the screen creation unit 56 of the server 10 is an example of the group information output unit and other group information output unit of the present invention.

[Student Device Configuration]
14 is a block diagram showing the hardware configuration of the student terminal 20. The student terminal 20 includes a communication unit 21, a control unit 22, a storage unit 23, a display unit 25, and an input unit 26. These components are connected to each other via a bus 29.

The communication unit 21 is a communication unit for communicating with the server 10 via the network 5. Specifically, the communication unit 21 sends answer information and target question information to the server 10, and receives prediction screens and recommendation screens from the server 10.

The control unit 22 includes a processing unit such as a CPU, MPU, or GPU, and performs various information processing, control processing, and the like related to the student terminal 20 by reading and executing programs stored in the memory unit 23. Note that while Figure 14 describes the control unit 22 as a single processor, it may also be a multi-processor.

The storage unit 23 includes memory elements such as RAM and ROM, and stores programs, data, etc. necessary for the control unit 22 to execute processing. The storage unit 23 also temporarily stores data, etc. necessary for the control unit 22 to execute arithmetic processing.

The display unit 25 is a liquid crystal display, organic EL display, or the like, and displays various information according to instructions from the control unit 22. The input unit 26 is an input device such as a mouse, keyboard, touch panel, or button, and outputs received operation information to the control unit 22.

[Screen creation process]
Next, a screen creation process will be described, which creates and outputs a prediction screen on which the subject can check the predicted correct answer rate and a recommendation screen on which the subject can refer to the answers to the target question by other students, based on the target question information entered by the subject. Fig. 15 is a flowchart of the screen creation process by the server 10. This process is realized by the server 10 executing a program prepared in advance.

The target student uses the student terminal 20 to send target question information, including a student ID and information about the target question, to the server 10. The server 10 acquires the target question information from the student terminal 20 (step S101). The server 10 then extracts from the learning history DB 43 the students who worked on the target question and answer information related to each student's answer to the target question (step S102). The server 10 then classifies each student's answer to the target question into multiple groups based on the pattern of answer content, based on the extracted answer information (step S103).

The server 10 calculates the accuracy rate of the answers belonging to each group for each group based on the answer information extracted by the extraction unit 52 (step S104). Then, the server 10 identifies the group to which the target answer belongs based on the target answer information (step S105). Furthermore, the server 10 creates a prediction screen that displays the predicted accuracy rate of the target answer and transmits it to the student terminal 20 used by the subject (step S106). By displaying the prediction screen received from the server 10 on the student terminal 20, the subject can check the predicted accuracy rate of their own answer or click a specified link to refer to example answers belonging to other groups.

The server 10 determines whether or not it has received information from the student terminal 20 that a specific link has been clicked by the subject (step S107). If it has not received information that a link has been clicked (step S107; No), the server 10 waits. On the other hand, if it has received information that a link has been clicked (step S107; Yes), the server 10 first identifies other groups corresponding to the clicked link. The server 10 then creates a recommendation screen including answers belonging to the identified other groups and transmits it to the student terminal 20 used by the subject (step S108). By displaying the recommendation screen received from the server 10 on the student terminal 20, the subject can easily check example answers that have answer patterns different from their own. This causes the server 10 to terminate the screen creation process.

Normally, when a problem requires calculation and there are insufficient intermediate steps or supplementary explanations in Japanese, the answer to the problem is inappropriate, which can result in a lower accuracy rate. The learning support system 100 of this embodiment can predict the accuracy rate of a target answer from the pattern of answer content, such as whether or not an intermediate step is included or whether or not supplementary explanations are provided in Japanese, for problems requiring calculation. In other words, it is possible to predict the accuracy rate from the pattern of answer content and evaluate whether or not the answer content is appropriate.

The learning assistance system 100 can also provide the subject with answers that belong to other groups. Specifically, the learning assistance system 100 can support the subject's learning by recommending answers that belong to other groups with a high rate of correct answers. This allows the subject to compare their own answers with the answers of other learners while referring to the correct answer rates of groups categorized by answer content patterns, and learn answering methods that are more likely to result in correct answers.

Second Embodiment
In the first embodiment, the server 10 calculates the accuracy rate of answers belonging to each group classified by answer content and displays it on the prediction screen, but the present invention is not limited to this, and it is also possible to calculate the academic ability level (hereinafter simply referred to as "academic ability") of students who worked on the answers belonging to each group and display it on the prediction screen.

First, the functional configuration of the server in the second embodiment will be described. Figure 16 is a block diagram showing the functional configuration of the server in the second embodiment. For the sake of simplicity, explanations similar to those in the first embodiment will be omitted, and only the LRT (Latent Rank Theory) evaluation value calculation unit 59 and the screen creation unit 56x will be described in detail.

The LRT evaluation value calculation unit 59 evaluates each student using the LRT based on the answer information extracted by the extraction unit 52, thereby calculating a numerical value for the academic ability level of students who worked on the answers belonging to each group. The LRT evaluation value can be, for example, the sum of the differences between the reference LRT and LRTs higher than the reference LRT. Specifically, if the reference LRT is "2" and the LRTs within the group are "3," "5," and "1," the LRT evaluation value is calculated as "4" using the formula (3 - 2) + (5 - 2). Alternatively, the LRT evaluation value can be set arbitrarily, such as the average LRT within the group or the total number of learners within the group with high LRTs exceeding a threshold. Figure 17 shows the LRT evaluation values for each group calculated by the LRT evaluation value calculation unit 59. In this embodiment, the LRT evaluation value is the average LRT within the group. The LRT evaluation value can visualize and provide information on which group the answers of students with high academic ability levels belong to.

When the screen creation unit 56x acquires the target question information, it references each DB and creates and outputs a prediction screen that checks the LRT evaluation value of the group to which the user's answer belongs.

Here, we will explain the prediction screen created by the screen creation unit 56x. The prediction screen is a screen that displays the LRT evaluation value of the group to which the target answer belongs and the LRT evaluation values of other groups, and is sent to the target student's student terminal 20. Figure 18(a) is an example of a prediction screen in the second embodiment. Note that, for the sake of convenience, explanations similar to those in the first embodiment will be omitted, and the group evaluation value item 90 and other group evaluation value item 91 will be explained in detail.

The group evaluation value item 90 displays the LRT evaluation value of the group to which the target answer belongs. The other group evaluation value item 91 displays the LRT evaluation values of other groups to which the target answer does not belong, and has a link that can be clicked to display answers belonging to other groups. Specifically, if the students' answers to the target question are classified into four groups, A to D, and the target answer belongs to group C, the group evaluation value item 90 displays group C's LRT evaluation value of "2.3" as "LRT evaluation value of same answer pattern: 2.3," as shown in Figure 18(a). Furthermore, the other group evaluation value item 91 displays group A's LRT evaluation value of "4.0" as "View example answers with an LRT evaluation value of 4.0," along with a link to display answers belonging to group A, as shown in Figure 18(a). The other group evaluation value item 91 also displays the LRT evaluation values for Group B and Group D, along with links to display the answers belonging to each group, such as "View example answers with an LRT evaluation value of 1.3" and "View example answers with an LRT evaluation value of 1.5."

When the screen creation unit 56x receives information from the student terminal 20 that a specific link has been clicked by the subject, it creates and outputs a recommendation screen that includes answers belonging to the group corresponding to the link.

Here, we will explain the recommendation screen created by the screen creation unit 56x. The recommendation screen is a screen that displays answers belonging to other groups to which the target answer does not belong, and is sent to the student terminal 20 of the target student. By displaying the recommendation screen received from the server 10x using the student terminal 20, the target student can refer to answers from groups with answer content patterns different from their own. Figure 18(b) is an example of a recommendation screen. Note that, for the sake of convenience, explanations similar to those in the first embodiment will be omitted, and the answer column 86 and group evaluation value item 92 will be explained in detail.

The answer column 86 displays answers belonging to other groups to which the target answer does not belong. Specifically, the screen creation unit 56x identifies the group corresponding to the link clicked by the target user in the other group evaluation value item 91 on the prediction screen and creates a recommendation screen to display answers belonging to that group. For example, if the target user uses the student device 20 to click "View example answers with an LRT evaluation value of 4.0" in the other group evaluation value item 91 on the prediction screen as shown in FIG. 18(a), the screen creation unit 56x identifies Group A, which has an LRT evaluation value of "4.0," and creates a recommendation screen to display answers belonging to Group A in the answer column 86. The answer column 86 can display multiple answers belonging to Group A using scroll buttons, next buttons, etc. The display order of the multiple answers belonging to Group A can be set arbitrarily, such as alphabetical order by student name or by student academic level. The group evaluation value item 92 displays the LRT evaluation value of the group to which the answer displayed in the answer column 86 belongs.

In this embodiment, the other group evaluation value item 91 on the prediction screen displays the LRT evaluation values and links of all other groups to which the target answer does not belong. However, the present invention is not limited to this, and it is also possible to display the LRT evaluation values and links of only other groups with LRT evaluation values higher than the LRT evaluation value of the group to which the target answer belongs, or only the other group with the highest LRT evaluation value. Specifically, the other group evaluation value item 91 may display the LRT evaluation value and link of the other group with the highest LRT evaluation value, along with the sentence recommended as learning support.

In addition, the screen creation unit 56x may not provide the other group evaluation value item 91 on the prediction screen, but may instead display the LRT evaluation value and a link in the group evaluation value item 90, and create a recommendation screen that allows users to refer to other answers that belong to the same group as the target answer when the link is clicked.

In the above configuration, the LRT evaluation value calculation unit 59 of the server 10x is an example of an academic ability calculation unit of the present invention. Also, the LRT evaluation value is an example of an academic ability evaluation value of the present invention.

[Screen creation process]
Next, a screen creation process will be described in which a prediction screen on which the subject can check the LRT evaluation value of the group to which the subject's answer belongs and a recommendation screen on which the subject can refer to the answers of other students to the target question are created and output based on the target question information entered by the subject. Figure 19 is a flowchart of the screen creation process by the server 10x. This process is realized by the server 10x executing a program prepared in advance.

The target student uses the student terminal 20 to send target question information, including a student ID and information about the target question, to the server 10x. The server 10x acquires the target question information from the student terminal 20 (step S201). The server 10x then extracts from the learning history DB 43 the students who worked on the target question and answer information related to each student's answer to the target question (step S202). The server 10x then classifies each student's answer to the target question into multiple groups based on the pattern of answer content, based on the extracted answer information (step S203).

Based on the answer information extracted by the extraction unit 52, the server 10x calculates the LRT evaluation value for each group of students who worked on the answers belonging to each group (step S204). Then, based on the target answer information, the server 10x identifies the group to which the target answer belongs (step S205). Furthermore, the server 10x creates a prediction screen that displays the LRT evaluation value of the group to which the target answer belongs, and sends it to the student terminal 20 used by the subject (step S206). By having the student terminal 20 display the prediction screen received from the server 10x, the subject can check the LRT evaluation value of the group to which their answer belongs, or click a specified link to refer to example answers belonging to other groups.

The server 10x determines whether it has received information from the student terminal 20 indicating that a specific link has been clicked by the subject (step S207). If it has not received information indicating that a link has been clicked (step S207; No), the server 10x waits. On the other hand, if it has received information indicating that a link has been clicked (step S207; Yes), the server 10x first identifies the other group corresponding to the clicked link. The server 10x then creates a recommendation screen including answers belonging to the identified other group and transmits it to the student terminal 20 used by the subject (step S208). By having the student terminal 20 display the recommendation screen received from the server 10x, the subject can easily check example answers that have answer patterns different from their own. This causes the server 10x to end the screen creation process.

The learning support system 100x of this embodiment can identify the group to which a target answer belongs based on the pattern of answer content, such as whether or not intermediate steps are included and whether or not supplementary explanations are provided in Japanese, for problems requiring calculations, and can provide the academic level of students who have worked on answers belonging to the same group. This allows the target person to check their academic level according to the content of their own answer.

The learning support system 100x can also provide the subject with answers that belong to other groups. Specifically, the learning support system 100x can support the subject's learning by recommending answers that belong to other groups with the highest LRT evaluation value. This allows the subject to compare their own answers with those of other learners while referring to the LRT evaluation values of groups categorized by answer content patterns, and learn the answering methods of students with high academic levels.

<Modification>
Next, modifications of the first and second embodiments will be described. The following modifications can be applied to the first and second embodiments in appropriate combinations.

(First Modification)
In the first and second embodiments described above, the extraction unit 52 extracts the students who worked on the target questions and the answer information of each student to the target questions from the learning history DB 43. Furthermore, the group classification unit 53 classifies the answers to the target questions of each student into groups subdivided according to the patterns of the answer content, based on the answer information extracted by the extraction unit 52.

However, the present invention is not limited to this. If only a few students have worked on the target question, the extraction unit 52 may further extract from the learning history DB 43 students who have worked on similar questions similar to the target question, and answer information for each student to the similar questions. In this case, the group classification unit 53 classifies each student's answers to the target question and similar questions into groups subdivided by answer content pattern, based on the answer information extracted by the extraction unit 52. Also, in this case, the answer column 86 on the recommendation screen preferentially displays answers to target questions belonging to other groups. If there is no answer to the target question, the answer column 86 on the recommendation screen displays answers to similar questions belonging to other groups.

Specifically, similar questions can be arbitrarily set to questions with the same learning elements as the target question, questions with the same learning elements and difficulty as the target question, etc. For example, if similar questions are set to questions with the same learning elements as the target question, the extraction unit 52 references the question DB 42 and identifies questions with the same learning elements as the target question as similar questions. The extraction unit 52 then extracts, from the learning history DB 43, information on the students who worked on the target question and their answers to the target question, and information on the students who worked on similar questions and their answers to the similar questions.

Note that, while this modified example uses learning elements to identify similar questions, the present invention is not limited to this, and any standard, such as the curriculum code established by the Ministry of Education, Culture, Sports, Science and Technology, may be used. The extraction unit 52 of this modified example is an example of a similar question identification unit of the present invention. The question DB 42 is an example of a question storage unit of the present invention.

(Second Modification)
In the first and second embodiments described above, the prediction screen and the recommendation screen are displayed as two screens in separate windows, but the present invention is not limited to this, and the screen layout can be set as desired, such as displaying the accuracy rate or LRT evaluation value of the group to which the target answer belongs and answers belonging to other groups on a single screen. For example, the screen creation unit 56 may create a screen that displays the target answer and answers belonging to other groups side by side to make them easier to compare, or may create a screen that also displays a model answer.

(Third Modification)
Furthermore, in the first and second embodiments described above, students use the student terminal 20 to send target question information to the server 10 and view the prediction screen and recommendation screen obtained from the server 10. However, the present invention is not limited to this. A teacher may use the teacher terminal 30 to send target question information, including the student ID of the target student and information about the target question, to the server 10 and view the prediction screen and recommendation screen obtained from the server 10. In this case, the server 10 creates a prediction screen based on the target question information received from the teacher terminal 30 and sends it to the teacher terminal 30. Furthermore, when the server 10 receives information from the teacher terminal 30 that the teacher has clicked a specific link on the prediction screen, it creates a recommendation screen including answers belonging to the group corresponding to the link and sends it to the teacher terminal 30.

(Fourth Modification)
In the first and second embodiments described above, the teacher uses the teacher terminal 30, but the present invention is not limited to this, and the teacher may use a teacher terminal 95 that has the functions of the server 10. In this case, the student terminals 20 are used by students, their parents, etc. The teacher terminal 95, like the server 10, is, for example, a personal computer or a general-purpose tablet PC (personal computer).

Figure 20 shows an example configuration of a learning support system 200 in this case. As shown, the learning support system 200 is configured to enable communication between a teacher terminal 95 and multiple student terminals 20 via a network 5. The teacher terminal 95 is connected to a learning element DB 96, a question DB 97, and a learning history DB 98, and can execute the screen creation process previously performed by the server 10 to classify students' answers to target questions by answer content pattern and create each screen. Therefore, the teacher terminal 95 can send the created prediction screens and recommendation screens to the student terminals 20 and display them on its own display. In this case, the teacher terminal 95 is an example of an information processing device of the present invention.

5 Network 10, 10x Server 20 Student terminal 30, 95 Teacher terminal 41, 96 Learning element DB
42, 97 Problem DB
43, 98 Learning history DB
51 Target question acquisition unit 52 Extraction unit 53 Group classification unit 54 Correct answer rate calculation unit 55 Group identification unit 56, 56x Screen creation unit 100, 100x, 200 Learning support system

Claims (10)

a terminal device used by a specific learner among the plurality of learners,
a learning history storage unit that stores answer information regarding answers to questions by a plurality of learners including the target learner;
a target question acquisition unit that acquires, from the terminal device, information about a target question that is a question designated by the subject;
an extraction unit that extracts answer information of the target question by a plurality of learners including the target person from the learning history storage unit;
a group classification unit that classifies the answers to the target questions by each of a plurality of learners including the target person into groups that are subdivided according to answer content patterns, based on the answer information extracted by the extraction unit;
a group identification unit that identifies a group to which a target answer, which is an answer to the target question by the subject, belongs based on target answer information, which is answer information to the target question by the subject;
An information processing device comprising: The information processing device according to claim 1, further comprising a group information output unit that outputs group information regarding the group to which the target answer belongs. a correct answer rate calculation unit that calculates a correct answer rate of answers that belong to the same group as the target answer based on the answer information extracted by the extraction unit;
The information processing device according to claim 2 , wherein the group information output unit outputs the correct answer rate as the group information. an academic ability calculation unit that calculates an evaluation value of the academic ability of a learner who has worked on an answer that belongs to the same group as the target answer, based on the answer information extracted by the extraction unit;
The information processing device according to claim 2 , wherein the group information output unit outputs the evaluation value of the academic ability as the group information. The information processing device according to claim 1 or 2, further comprising an other group information output unit that outputs other group information relating to other groups to which the target answer does not belong. a correct answer rate calculation unit that calculates the correct answer rate of the answers belonging to the other group based on the answer information extracted by the extraction unit;
The information processing device according to claim 5 , wherein the other group information output unit outputs the correct answer rate as the other group information. an academic ability calculation unit that calculates an evaluation value of the academic ability of the learners who have worked on the answers belonging to the other group based on the answer information extracted by the extraction unit;
The information processing device according to claim 5 or 6, wherein the other group information output unit outputs the evaluation value of the academic ability as the other group information. An information processing device according to any one of claims 5 to 7, wherein the other group information output unit outputs answers belonging to the other group as the other group information based on the answer information extracted by the extraction unit. The problem is linked to information of a learning element indicating a learning content, and a similar problem identification unit is provided that identifies a problem having the same learning element as a similar problem;
the extraction unit further extracts, from the learning history storage unit, learners who have tackled the similar problems and answer information relating to answers to the similar problems;
The information processing device according to claim 1, wherein the group classification unit classifies the answers to the target question and the similar questions by each of a plurality of learners, including the subject, into groups subdivided according to patterns of answer content based on the extracted answer information. A program executed by an information processing device having a computer,
the information processing device is communicably connected to a terminal device used by a target person who is a specific learner among a plurality of learners,
a learning history storage unit that stores answer information regarding answers to questions by a plurality of learners including the target learner;
a target question acquisition unit that acquires, from the terminal device, information about a target question that is a question designated by the subject;
an extraction unit that extracts answer information of the target question by a plurality of learners including the target person from the learning history storage unit;
a group classification unit that classifies the answers to the target questions by each of a plurality of learners including the subject person into groups subdivided according to answer content patterns based on the answer information extracted by the extraction unit;
a group identification unit that identifies a group to which a target answer, which is an answer to the target question by the subject, belongs, based on target answer information, which is answer information to the target question by the subject;
A program that causes the computer to function as a