Text reading comprehension training device, text reading comprehension training system, text reading comprehension training method, and text reading comprehension training program

Abstract

Improve reading comprehension skills. [Solution] The text reading training device comprises a text analysis unit that determines target text information for each target, including the theme and / or genre of the question text and the text structure of the question text; a pattern bank generation unit that generates a pattern bank in which target text information is registered for each target; a theme genre bank generation unit that generates a theme genre bank in which question generation text information, including the theme and / or genre of the question generation text and the text structure of the question generation text, is associated and registered; and a training unit that reads out target text information registered for a specific target from the pattern bank, reads out question generation text associated with question generation text information corresponding to the target text information from the theme genre bank, generates and outputs training questions using the read question generation text, and evaluates the user's answers.

Description

[Technical Field]

[0001] The present disclosure relates to a text reading training device, a text reading training system, a text reading training method, and a text reading training program for training a user's text reading ability. [Background technology]

[0002] A technology has been disclosed that enables individual test takers to efficiently improve their academic ability by outputting learning materials optimized for each test taker, such as in mock entrance exams (Patent Document 1). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Patent No. 7508162 Summary of the Invention [Problem to be solved by the invention]

[0004] The reading comprehension skills required for entrance exams are not only required for entrance exams, but also for various career stages such as qualification exams, employment exams, work assignments, and social issues. For this reason, it is desirable to develop reading comprehension skills that are applicable not only to entrance exams but also to various career stages such as qualification exams, employment exams, work assignments, and social issues. [Means for solving the problem]

[0005] A text reading training device according to an embodiment of the present disclosure includes: a text analysis unit that analyzes a question text, which is a text of a text comprehension question of a target, to determine target text information including a theme and / or genre of the question text and a text structure of the question text for each target; a pattern bank generating unit that generates, for each of the targets, a pattern bank in which the target sentence information is registered; a theme genre bank generation unit that generates a theme genre bank in which question generation texts, which are texts used to generate training reading comprehension questions, themes and / or genres of the question generation texts, and text structures of the question generation texts are associated and registered; reading out the target sentence information registered for a specific target from the pattern bank; reading out from the theme genre bank a question generation sentence associated with question generation sentence information corresponding to the target sentence information; The read sentences for generating questions are used to generate and output training questions, and the answers of the user are evaluated. Training Department and It is equipped with: [Effects of the Invention]

[0006] According to the present disclosure, it is possible to improve reading comprehension skills applicable to various career stages, such as entrance exams, qualification exams, employment exams, work assignments, and social assignments.

[0007] The effects described here are not necessarily limited to those described herein, and may be any of the effects described in this disclosure. [Brief explanation of the drawings]

[0008] [Figure 1] 1 illustrates a text reading comprehension training system according to an embodiment of the present disclosure. [Figure 2] 1 shows the functional configuration of a text reading comprehension training device. [Figure 3] 1 shows an information processing operation flow of a text reading training device. [Figure 4] 1 shows the functional configuration of the analysis unit. [Figure 5] An example of a thought code is shown below. [Figure 6] The functional configuration of the training section is shown below. [Figure 7] The functional configuration of the final evaluation unit is shown below. DETAILED DESCRIPTION OF THE INVENTION

[0009] Hereinafter, embodiments of the present disclosure will be described with reference to the drawings.

[0010] 1. Reading comprehension training system

[0011] FIG. 1 illustrates a text reading comprehension training system according to one embodiment of the present disclosure.

[0012] The text reading training system 1 includes a text reading training device 10 and an electronic device 20.

[0013] The electronic device 20 is a terminal device used by a user (question answerer). The electronic device 20 may typically be a general-purpose device such as a smartphone, tablet device, personal computer, or robot-type device, but may also be a device dedicated to a text reading comprehension training system. The electronic device 20 has a computer, an input device, and an output device. The input device is, for example, a microphone, a touch panel, a keyboard, or a mouse. The output device is, for example, a speaker or a display.

[0014] The text reading comprehension training device 10 is a server device capable of communicating with the electronic device 20 via a network N such as the Internet. The computer of the text reading comprehension training device 10 transmits questions to be answered by the user (text comprehension questions) and training questions to the electronic device 20, and causes the electronic device 20 to output them as voice, text, and / or images. The text comprehension questions are typically long reading comprehension questions in the user's native language or a foreign language. The computer of the text reading comprehension training device 10 receives answers input by the user via an input device as voice, text, and / or images. The text reading comprehension training device 10 evaluates the user's answers. Based on the user's answers, the text reading comprehension training device 10 determines the next question to be output to the user, and transmits that question to the electronic device 20 for output. By repeating this process, a dialogue between the text reading comprehension training device 10 and the user via the electronic device 20 is realized.

[0015] When the electronic device 20 outputs an image, the image may be an image representing a human being. When the electronic device 20 outputs a sound, the sound may be a sound representing the speech of the human being represented as an image. In this way, the reading comprehension training device 10 functions as an electronic tutor, mentor, or counselor via the electronic device 20.

[0016] 2. Functional configuration and information processing operation flow of the text reading training device

[0017] FIG. 2 shows the functional configuration of the text reading training device.

[0018] The text reading comprehension training device 10 operates as an analysis unit 100, a training unit 200, and a final evaluation unit 300 by having the CPU constituting the computer load a training program recorded in ROM into RAM and execute it. All or part of the analysis unit 100, training unit 200, and final evaluation unit 300 can utilize AI such as generation AI.

[0019] The training section 200 includes at least some (in this embodiment, all) of a pre-training section 211, a beginner's training section 212, an intermediate training section 213, an advanced training section 214, and a top-level training section 215.

[0020] FIG. 3 shows the information processing operation flow of the text reading training device.

[0021] 2-1.Analysis Department

[0022] FIG. 4 shows the functional configuration of the analysis unit.

[0023] The analysis unit 100 includes a sentence analysis unit 101 , a question analysis unit 102 , a pattern bank generation unit 103 , and a theme genre bank generation unit 104 .

[0024] The text analysis unit 101 analyzes the text (problem text) of the text comprehension question 30 of the target 111 to determine target text information 112 for each target 111, including the theme and / or genre 113 of the question text and the text structure 114 of the question text (step S110). The "text comprehension question 30 of the target 111" is, for example, a text comprehension question on a language subject in an entrance exam or a mock test. The "target 111" is, for example, a desired school for a junior high school entrance exam, a high school entrance exam, a university entrance exam, various qualifications for a qualification exam, a job type for an employment exam, a work assignment, a social issue, etc. The "theme and / or genre 113" is, for example, a theme and / or genre such as current affairs or book information.

[0025] The text analysis unit 101 analyzes the problem text using, for example, the Toulmin model. The Toulmin model is a stable model used in schools for research and essay writing. It analyzes whether the components of text structure, such as assertion, evidence, reasoning, counterargument, background, and qualification, are written and is suitable for identifying areas for improvement.

[0026] The question analysis unit 102 analyzes the questions in the reading comprehension questions 30 of the targets 111, and determines question information 122 including multiple thought code domains 123 that represent the characteristics of the questions and the difficulty level 124 of the questions for each target 111 (step S120). For example, the question analysis unit 102 analyzes the questions using the Toulmin model.

[0027] Figure 5 shows an example of a thought code.

[0028] An example of a thought code is a thought code 400 that represents the characteristics of questions in an entrance exam or a mock exam. The thought code 400 includes multiple (nine in this example) thought code domains A1, A2, A3, B1, B2, B3, C1, C2, and C3.

[0029] The multiple thought code domains 123 (A1, A2, A3, B1, B2, B3, C1, C2, C3) are arranged in a 3x3 matrix, consisting of a first axis 411, which is the horizontal axis, and a second axis 412, which intersects (orthogonal in this example) with the first axis 411. The first axis 411 represents the level of thinking (three levels in this example). The second axis 412 represents the level of knowledge (three levels in this example).

[0030] The first axis 411 indicates that the thinking level increases in three stages, with the thinking code domain A1 as the origin, as one moves away from the thinking code domain A1 (progressing to thinking code domains B1 and C1). Specifically, "A" represents the perspective of "whether one has knowledge and understands," "B" represents "whether one can apply it and think and express it logically," and "C" represents "whether one can think and express it critically and creatively."

[0031] The second axis 412 indicates that the knowledge level increases in three stages as one moves away from the thought code domain A1 (progressing to the thought code domains A2, A3, and so on). Specifically, "1" represents a scale of "simple knowledge level problems," "2" represents "complex knowledge level problems," and "3" represents "transformative (higher level) knowledge level problems."

[0032] Thinking Code Domain A1 represents the ability to "knowledge and understand simple knowledge-level problems," specifically related to general grading criteria. Thinking Code Domain A2 represents the ability to "knowledge and understand complex knowledge-level problems," specifically related to clarity of argument. Thinking Code Domain A3 represents the ability to "knowledge and understand transformative (higher-level) knowledge-level problems," specifically related to the interestingness of the ideas.

[0033] Thinking Code Domain B1 represents "whether it can be applied to problems at a simple knowledge level and whether it can be thought of and expressed logically," specifically relating to specific examples, data, and episodes. Thinking Code Domain B2 represents "whether it can be applied to problems at a complex knowledge level and whether it can be thought of and expressed logically," specifically relating to the connection between claims and reasons, counterarguments to claims, and reasons that support those counterarguments. Thinking Code Domain B3 represents "whether it can be applied to problems at a transformative (higher) knowledge level and whether it can be thought of and expressed logically," specifically relating to multifaceted thinking and the degree to which the five eyes (bird's eyes, insect's eyes, fish's eyes, bat's eyes, and mind's eyes) are utilized.

[0034] Thinking code domain C1 represents "the ability to think and express critically and creatively on simple knowledge-level problems," specifically, relating to clarifying the conditions of how important something is to whom. Thinking code domain C2 represents "the ability to think and express critically and creatively on complex knowledge-level problems," specifically, relating to the historical background and story that supports the reasons. Thinking code domain C3 represents "the ability to think and express critically and creatively on transformative (higher-level) knowledge-level problems," specifically, relating to original ideas.

[0035] The pattern bank generation unit 103 generates a pattern bank 130 in which target sentence information 112 (theme and / or genre 113, sentence structure 114) and question information 122 (thought code domain 123, difficulty level 124) are registered for each target 111 (step S130).

[0036] The theme genre bank generation unit 104 generates a theme genre bank 140 in which text used to generate training text comprehension questions (text for question generation 141) is associated with and registered as text for question generation information 142 including the theme and / or genre 143 of the text for question generation 141 and the text structure 144 of the text for question generation 141 (step S140).

[0037] 2-2. Training Department

[0038] FIG. 6 shows the functional configuration of the training section.

[0039] The training unit 200 includes a target sentence information determination unit 201, a training question generation sentence determination unit 202, a question information determination unit 203, a level determination unit 204, and a training execution unit 205. Specifically, the training unit 200 executes the sentence comprehension training as follows (step S150).

[0040] The target sentence information determination unit 201 reads out the target sentence information 112 (theme and / or genre 113, sentence structure 114) registered for a specific target 111 (for example, a specific desired school) from the pattern bank 130 (step S151).

[0041] The training question generation sentence determination unit 202 reads out question generation sentences 141 associated with question generation sentence information 142 (theme and / or genre 143, sentence structure 144) that corresponds (typically matches) to the target sentence information 112 (theme and / or genre 113, sentence structure 114) from the theme genre bank 140 (step S152).

[0042] The question information determination unit 203 reads out the question information 122 (thought code domain 123, difficulty level 124) registered for the specific target 111 from the pattern bank 130 (step S153).

[0043] The level determination unit 204 determines the target level that the user should achieve for the target 111 from among multiple levels based on the read question information 122 (thought code domain 123, difficulty level 124) (step S154). The "multiple levels" refer to multiple consecutive levels from the basic level to the most advanced level, specifically the basic level, beginner level, intermediate level, advanced level, and most advanced level.

[0044] The training execution unit 205 generates training questions to achieve the target level using the read question generation text 141 based on the read question information 122 (thinking code domain 123, difficulty level 124), outputs the questions to the electronic device 20, and evaluates the user's answers input from the electronic device 20 (step S155).

[0045] According to this embodiment, the training execution unit 205 can generate training questions using question-generating sentences 141 associated with question-generating sentence information 142 (theme and / or genre 143, sentence structure 144) that corresponds to (typically matches) target sentence information 112 (theme and / or genre 113, sentence structure 114) registered for a specific target 111 (e.g., a specific desired school). This generates training questions using sentences that match the target, rather than sentences unrelated to the target, allowing the user to perform appropriate training to solve the target's text comprehension problems. Furthermore, the training execution unit 205 can generate training questions based on question information 122 (thought code domain 123, difficulty level 124) registered for the specific target 111 (e.g., a specific desired school). This generates training questions using a question format that matches the target, rather than a question format unrelated to the target, allowing the user to perform more appropriate training to solve the target's text comprehension problems.

[0046] The training execution unit 205 generates and outputs training questions of one or more consecutive levels from the basic level to the target level, in ascending order without interruption from the basic level to the target level, for training at each of a plurality of consecutive levels from the basic level to the most advanced level (basic level, beginner level, intermediate level, advanced level, most advanced level), and evaluates the user's answers. For example, when the target level is the advanced level, the training execution unit 205 does not implement only the advanced level training, but implements the basic level training, beginner level training, intermediate level training, and advanced level training in this order (ascending order) without interruption.

[0047] According to this embodiment, the level determination unit 204 determines the target level that the user should achieve for the target 111 from a plurality of levels (basic level, beginner level, intermediate level, advanced level, and highest level). The training execution unit 205 generates and outputs training questions in ascending order from the basic level to the target level without interruption, and evaluates the user's answers. By performing training in stages in this manner, the user can perform appropriate training to achieve the target level.

[0048] The training execution unit 205 includes a pre-training unit 211, a beginner's training unit 212, an intermediate training unit 213, an advanced training unit 214, and a top-level training unit 215. The pre-training unit 211 executes training for a user to reach the basic level. The beginner's training unit 212 executes training for a user who has achieved the basic level to reach the next level, the beginner's level. The intermediate training unit 213 executes training for a user who has achieved the beginner's level to reach the next level, the intermediate level. The advanced training unit 214 executes training for a user who has achieved the intermediate level to reach the next level, the advanced level. The top-level training unit 215 executes training for a user who has achieved the advanced level to reach the next level, the top-level level. The pre-training unit 211, the beginner's training unit 212, the intermediate training unit 213, the advanced training unit 214, and the top-level training unit 215 each generate different types of training questions. Specific explanations are provided below.

[0049] The pre-training unit 211 generates and outputs pre-training questions that analyze the question generation sentence 141 for each component, and evaluates the user's answers so that the user can achieve the basic level. For example, the pre-training unit 211 outputs questions that check the sentence for each component of a sentence model (two-min model). The pre-training unit 211 accepts answers entered by the user in writing (or voice) and interacts with the user until the answers match the components of the sentence model output by the generation AI as a model answer. The pre-training unit 211 displays the matching rate up to a predetermined number of times (e.g., the third time), and determines that the basic level has been achieved if there is a match of 80% or more. If there is no match of 80% or more, from the predetermined number of times (e.g., the fourth time) onwards, the pre-training unit 211 provides feedback to the user on the components of the sentence model output by the generation AI as a model answer, and the user answers accordingly.

[0050] The beginner's training unit 212 generates and outputs beginner's training questions, which are questions to create summaries of the question generation sentences 141, so that a user who has achieved the basic level can reach the beginner's level, and evaluates the user's answers. For example, the beginner's training unit 212 generates beginner's training questions, which are questions to create summaries of sentences by gradually increasing the number of characters (50 characters, 100 characters, 200 characters, etc.). The beginner's training unit 212 accepts answers entered by the user in writing (or voice) and interacts with the user until the answers match the summaries output by the generation AI as model answers. The beginner's training unit 212 displays the matching rate up to a predetermined number of times (e.g., the third time), and determines that the beginner's level has been achieved if there is an 80% or higher match. If there is no match of 80% or more, the beginner's training unit 212 provides feedback to the user on the summaries output as model answers from the predetermined number of times (e.g., the fourth time) onwards, and the user answers accordingly.

[0051] The intermediate training unit 213 generates and outputs intermediate training questions, which are questions that a user who has achieved the beginner level fills in blanks provided in the question generation text 141 in order to achieve the intermediate level, and evaluates the user's answers. For example, the intermediate training unit 213 generates intermediate training questions by providing blanks for each paragraph of a text and prompting the user to fill in the blanks. The blanks are used to confirm the components of a text model (Two-Min model). The intermediate training unit 213 accepts answers entered by the user in writing (or speech) and interacts with the user until the answers match the words and phrases output by the generation AI as model answers. The intermediate training unit 213 displays the matching rate up to a predetermined number of times (e.g., the third time), and determines that the intermediate level has been achieved if there is an 80% or higher match. If there is no 80% or higher match, the intermediate training unit 213 provides feedback to the user of the words and phrases output as model answers from the predetermined number of times (e.g., the fourth time) onwards, and the user answers accordingly.

[0052] The advanced training unit 214 generates and outputs advanced training questions that analyze the structure (sentence components) of the question generation sentence 141 so that a user who has achieved the intermediate level can achieve the advanced level, and evaluates the user's answers. For example, the advanced training unit 214 generates advanced training questions in which the user analyzes the sentence structure of a reading comprehension sentence. The advanced training unit 214 uses a generation AI to check whether the sentence components of the user's answer match a sentence model (Two-Min model). The advanced training unit 214 accepts an answer entered by the user in writing (or voice) and interacts with the user until it matches the structure (sentence components) output by the generation AI as a model answer. The advanced training unit 214 displays the matching rate up to a predetermined number of times (e.g., the third time), and determines that the advanced level has been achieved if there is an 80% or higher match. If there is no 80% or higher match, the advanced training unit 214 provides feedback to the user on the structure (sentence components) output as a model answer from the predetermined number of times (e.g., the fourth time) onwards, and the user answers accordingly.

[0053] The highest level training unit 215 generates and outputs highest level training questions that generate questions for the question generation text 141 so that a user who has achieved the advanced level can achieve the highest level, and evaluates the user's answers. For example, the highest level training unit 215 creates questions for text comprehension. For example, the perspective of the questions is compared with the Goodman model (five perspectives: "deletion and insertion," "ordering," "weighting," "decomposition and integration," and "transformation"). The user creates a predetermined number of questions (e.g., 10 questions). The highest level training unit 215 accepts answers entered by the user in writing (or voice) and uses a generation AI to interact until the answers match the Goodman model. The highest level training unit 215 displays the matching rate up to a predetermined number of times (e.g., the fourth time), and determines that the highest level has been achieved if the answers match 80% with the Goodman model. If the Goodman model does not match 80% or more, after a certain number of times (for example, the fifth time), the user is given feedback on questions that match the Goodman model output as model answers, and the user answers accordingly.

[0054] The Goodman Model is said to be a perspective of scientific and mathematical thinking that integrates the humanities and sciences, and is a sense that can be used when considering mathematics for prestigious national universities. It is a conceptual lens for generating ideas for art thinking and design thinking.

[0055] 2-3. Final evaluation section

[0056] FIG. 7 shows the functional configuration of the final evaluation unit.

[0057] The final evaluation unit 300 includes a target information determination unit 301, a sentence determination unit 302 for generating final evaluation questions, and an evaluation unit 303. The final evaluation unit 300 generates and outputs sentence comprehension questions and evaluates the user's answers (step S160), specifically as follows:

[0058] The target information determination unit 301 reads out the target sentence information 112 (theme and / or genre 113, sentence structure 114) and question information 122 (thinking code domain 123, difficulty level 124) registered for a specific target 111 (e.g., a desired school) from the pattern bank 130 (step S161).

[0059] The final evaluation question generation sentence determination unit 302 reads out question generation sentence 141 associated with question generation sentence information 142 (theme and / or genre 143, sentence structure 144) corresponding to target sentence information 112 (theme and / or genre 113, sentence structure 114) from the theme genre bank 140 (step S162).

[0060] The evaluation unit 303 generates and outputs a text comprehension question using the read-out question generation text 141 based on the read-out question information 122, and evaluates the user's answer (step S163). For example, the evaluation unit 303 generates five long-passage questions.

[0061] The user solves the questions, and the evaluation unit 303 determines that the user has passed if all questions (five questions in this example) are answered correctly at a predetermined percentage or more. The evaluation unit 303 continues to present new long-passage questions for the number of questions for which the correct answer rate did not reach the predetermined percentage. The evaluation unit 303 repeats this process until all questions have a correct answer rate of at least the predetermined percentage. The predetermined percentage may be a fixed value (e.g., 70 to 80%) or may vary depending on the question. If it varies depending on the question, the evaluation unit 303 generates an expected correct answer rate for the text comprehension questions based at least on the multiple thinking code domains 123 included in the read question information 122, and evaluates the user's answer based on the expected correct answer rate (step S164).

[0062] According to this embodiment, the evaluation unit 303 generates a predicted correct answer rate based on the thought code domain 123, and evaluates the user's answer based on the predicted correct answer rate. Since the thought code 400 expresses the high or low of the thinking level and the high or low of the knowledge level, generating a predicted correct answer rate based on the thought code domain 123 increases the possibility of accurately generating a predicted correct answer rate. As a result, the possibility of more accurately evaluating the user's answer also increases.

[0063] 3. Applications of this embodiment

[0064] The reading comprehension ability required for entrance exams is required not only for entrance exams but also for various career stages such as qualification exams, employment exams, work assignments, and social issues. According to this embodiment, it is possible to improve reading comprehension ability applicable not only to entrance exams but also to various career stages such as qualification exams, employment exams, work assignments, and social issues.

[0065] According to this embodiment, by repeatedly achieving successive levels in stages, the text reading training device 1 of this embodiment can create a portfolio of the changes in the user's training ability, making clear the degree of growth in text reading skills for each life goal.

[0066] Examples of targets include junior high school entrance exam targets (desired schools), high school entrance exam targets, university entrance exam targets, various qualification targets, work problem-solving targets, and social problem-solving targets. Each target requires a path to career realization, such as the following. For example, the career realization goal may be passing the desired school for an entrance exam, obtaining a qualification for a qualification exam, joining a company for an employment exam, problem-solving for a work issue, and contributing to society and problem-solving for a social issue. The text reading comprehension training device 1 of this embodiment makes it possible to achieve a path to career realization by repeating training feedback.

[0067] Other application examples include: Proximal development zone to model sentence for passing: The sentence reading comprehension training device 1 may obtain a re-scoring of the respondent's answers to the training questions and further suggest advice that takes into account the respondent's intentions via the web and app.

[0068] Strategic learning for passing: The reading comprehension training device 1 may timely determine whether the chances of passing the desired school have increased. This can increase motivation through realistic reflection.

[0069] Visualizing the potential of students' thinking and expressive abilities after enrollment: By comparing the population correct answer rate with the individual correct answer rate of each respondent, the reading comprehension training device 1 can determine the level of effort required to improve students' thinking and expressive abilities after enrollment. Based on this, it can provide feedback on the order in which thinking code domains students should work on (starting with areas of strength).

[0070] Competency (skill) evaluation system (sold separately) = Automatically delivers questions to train the ability to think and express oneself in writing for career advancement: Questions to train each area are provided according to the order of areas to be trained, and the text reading training device 1 can be applied to individually optimize this.

[0071] Application to essay preparation for university entrance examinations: The reading comprehension training device 1 may be applied to individual optimization of essays for university entrance examinations.

[0072] Development into a talent discovery system (talent evaluation from writing): The reading comprehension training device 1 may be applied to grading essays for working adults. It may also be used in individual optimization systems for test takers, high school students, university students, and teachers.

[0073] Although the embodiments and modified examples of the present technology have been described above, the present technology is not limited to the above-described embodiments, and it goes without saying that various modifications can be made within the scope of the gist of the present technology. [Explanation of symbols]

[0074] 1: Text reading comprehension training system, 10: Text reading comprehension training device, 100: Analysis unit, 101: Text analysis unit, 102: Question analysis unit, 103: Pattern bank generation unit, 104: Theme genre bank generation unit, 111: Target, 112: Target text information, 113: Genre, 114: Text structure, 122: Question information, 123: Thinking code domain, 124: Difficulty level, 130: Pattern bank, 140: Theme genre bank, 141: Text for question generation, 142: Text information for question generation, 143: Genre, 144: Text structure, 20: Electronic device, 2 00: Training section, 201: Target text information determination section, 202: Text determination section for generating training questions, 203: Question information determination section, 204: Level determination section, 205: Training execution section, 211: Pre-training section, 212: Beginner training section, 213: Intermediate training section, 214: Advanced training section, 215: Highest level training section, 30: Text comprehension questions, 300: Final evaluation section, 301: Target information determination section, 302: Text determination section for generating final evaluation questions, 303: Evaluation section, 400: Thinking code, 411: First axis, 412: Second axis.

Claims

1. a text analysis unit that analyzes a question text, which is a text of a text comprehension question of a target, to determine target text information including a theme and / or genre of the question text and a text structure of the question text for each target; a question analysis unit that analyzes the questions of the reading comprehension questions of the target to determine, for each target, question information including a plurality of thought code domains that represent the characteristics of the questions and the difficulty level of the questions; a pattern bank generating unit that generates a pattern bank in which the target sentence information and the question information are registered for each of the targets; a theme genre bank generation unit that generates a theme genre bank in which question generation texts, which are texts used to generate training reading comprehension questions, themes and / or genres of the question generation texts, and text structure of the question generation texts are associated and registered; reading out the target sentence information registered for a specific target from the pattern bank; Read out the question information registered for the specific target from the pattern bank; determining a target level that the user should achieve for the target from a plurality of levels based on the read question information; reading out from the theme genre bank a question generation sentence associated with question generation sentence information corresponding to the target sentence information; Based on the read question information, and using the read question generation sentence, a training question for achieving the target level is generated and output, and the user's answer is evaluated. Training Department and Equipped with The plurality of thought code domains are in a matrix form consisting of a first axis representing the level of thought and a second axis intersecting the first axis representing the level of knowledge. A reading comprehension training device.

2. 2. The text reading training device according to claim 1, The training section the training system has at least a part including a pre-training section for a user to reach the basic level, a beginner's training section for a user who has achieved the basic level to reach the next level, which is a beginner's level, an intermediate training section for a user who has achieved the beginner's level to reach the next level, which is an intermediate level, an advanced training section for a user who has achieved the intermediate level to reach the next level, which is an advanced level, and a highest level training section for a user who has achieved the advanced level to reach the next level, which is the highest level, generating and outputting training questions of one or more successive levels from the basic level to the target level in ascending order without interruption from the basic level to the target level, and evaluating the user's answers; A reading comprehension training device.

3. 3. The text reading training device according to claim 2, The pre-training unit generates and outputs pre-training questions that analyze sentences for question generation for each component so that the user can reach the basic level, and evaluates the user's answers. A reading comprehension training device.

4. 3. The text reading training device according to claim 2, The beginner's training unit generates and outputs beginner's training questions, which are questions for creating summaries of question generation texts, so that a user who has achieved the basic level can reach the beginner's level, and evaluates the user's answers. A reading comprehension training device.

5. 3. The text reading training device according to claim 2, The intermediate training unit generates and outputs intermediate training questions, which are questions that a user who has achieved the beginner level fills in blanks provided in a question generation sentence in order to reach the intermediate level, and evaluates the user's answers. A reading comprehension training device.

6. 3. The text reading training device according to claim 2, The advanced training unit generates and outputs advanced training questions that are questions for analyzing the structure of question generation sentences so that a user who has achieved the intermediate level can reach the advanced level, and evaluates the user's answers. A reading comprehension training device.

7. 3. The text reading training device according to claim 2, The highest level training unit generates and outputs highest level training questions that generate questions for question generation sentences so that a user who has achieved the advanced level can achieve the highest level, and evaluates the user's answers. A reading comprehension training device.

8. 2. The text reading training device according to claim 1, reading out the target sentence information and the question information registered for a specific target from the pattern bank; reading out from the theme genre bank a question generation sentence associated with question generation sentence information corresponding to the target sentence information; A text comprehension question is generated and output based on the read question information using the read text for question generation, and the user's answer is evaluated. Final Evaluation Section The text reading comprehension training device further comprises:

9. 9. The text reading training device according to claim 8, The final evaluation unit generates a predicted correct answer rate for the reading comprehension question based at least on the plurality of thought code domains included in the read question information, and evaluates the user's answer based on the predicted correct answer rate. A reading comprehension training device.

10. A text reading comprehension training device; an electronic device capable of communicating with the text reading training device; The text reading training device includes: a text analysis unit that analyzes a question text, which is a text of a text comprehension question of a target, to determine target text information including a theme and / or genre of the question text and a text structure of the question text for each target; a question analysis unit that analyzes the questions of the reading comprehension questions of the target to determine, for each target, question information including a plurality of thought code domains that represent the characteristics of the questions and the difficulty level of the questions; a pattern bank generating unit that generates a pattern bank in which the target sentence information and the question information are registered for each of the targets; a theme genre bank generation unit that generates a theme genre bank in which question generation texts, which are texts used to generate training reading comprehension questions, themes and / or genres of the question generation texts, and text structure of the question generation texts are associated and registered; reading out the target sentence information registered for a specific target from the pattern bank; Read out the question information registered for the specific target from the pattern bank; determining a target level that the user should achieve for the target from a plurality of levels based on the read question information; reading out from the theme genre bank a question generation sentence associated with question generation sentence information corresponding to the target sentence information; Based on the read question information and using the read question generation sentence, a training question for achieving the target level is generated and output to the electronic device, and the user's answers input from the electronic device are evaluated. Training Department and and The plurality of thought code domains are in a matrix form consisting of a first axis representing the level of thought and a second axis intersecting the first axis representing the level of knowledge. Reading comprehension training system.

11. The control circuit of the text reading training device executes the text reading training program, By analyzing a problem sentence, which is a sentence of a reading comprehension question of a target, determining target sentence information for each target, including a theme and / or genre of the problem sentence and a sentence structure of the problem sentence; By analyzing the questions of the reading comprehension questions of the target, question information including a plurality of thought code domains representing the characteristics of the questions and the difficulty level of the questions is determined for each of the targets; generating a pattern bank in which the target sentence information and the question information are registered for each of the targets; generating a theme genre bank in which question generation texts, which are texts used to generate training reading comprehension questions, and question generation text information including the themes and / or genres of the question generation texts and the text structures of the question generation texts are associated and registered; reading out the target sentence information registered for a specific target from the pattern bank; Read out the question information registered for the specific target from the pattern bank; determining a target level that the user should achieve for the target from a plurality of levels based on the read question information; reading out from the theme genre bank a question generation sentence associated with question generation sentence information corresponding to the target sentence information; generating and outputting training questions for achieving the target level based on the retrieved question information and using the retrieved question generation sentences, and evaluating the user's answers; The plurality of thought code domains are in a matrix form consisting of a first axis representing the level of thought and a second axis intersecting the first axis representing the level of knowledge. Reading comprehension training method.

12. The control circuit of the text reading training device is a text analysis unit that analyzes a question text, which is a text of a text comprehension question of a target, to determine target text information including a theme and / or genre of the question text and a text structure of the question text for each target; a question analysis unit that analyzes the questions of the reading comprehension questions of the target to determine, for each target, question information including a plurality of thought code domains that represent the characteristics of the questions and the difficulty level of the questions; a pattern bank generating unit that generates a pattern bank in which the target sentence information and the question information are registered for each of the targets; a theme genre bank generation unit that generates a theme genre bank in which question generation texts, which are texts used to generate training reading comprehension questions, themes and / or genres of the question generation texts, and text structure of the question generation texts are associated and registered; reading out the target sentence information registered for a specific target from the pattern bank; Read out the question information registered for the specific target from the pattern bank; determining a target level that the user should achieve for the target from a plurality of levels based on the read question information; reading out from the theme genre bank a question generation sentence associated with question generation sentence information corresponding to the target sentence information; Based on the read question information, and using the read question generation sentence, a training question for achieving the target level is generated and output, and the user's answer is evaluated. Training Department A text reading comprehension training program that operates as follows: The plurality of thought code domains are in a matrix form consisting of a first axis representing the level of thought and a second axis intersecting the first axis representing the level of knowledge. Reading comprehension training program.

Images