JP2015106257A - Learning support device, learning support method and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To propose improvement of learning to a learning person so as to sufficiently improve learning effects.SOLUTION: A learning support device 1 includes a difference acquisition unit 30 and an improvement proposal unit 40. On the basis of a determination tree that is a learning model for a plurality of learning persons and learning log data for an improvement proposal object person, the difference acquisition unit 30 calculates a specific high achiever group formed from persons with higher learning results than the improvement proposal object person and with learning situations close to the improvement proposal object person, and calculates a difference of the learning log data between the improvement proposal object person and the specific high achiever group. On the basis of the difference calculated by the difference acquisition unit 30, the improvement proposal unit 40 proposes improvement to the improvement proposal object person.

Description

  The present invention relates to a learning support device, a learning support method, and a program.

  2. Description of the Related Art Conventionally, there has been proposed a technique for making an improvement proposal regarding learning, such as a learning method, to a learner who uses e-learning or online learning (see Non-Patent Document 1, for example).

  In such a technique, first, learning log data, which is information when learning is performed, is collected for one or more learners who use e-learning or online learning. Next, a learner's model is constructed by machine learning using the collected learning log data. Next, based on the model that has been built and the learning log data of the improvement proposal target person who makes improvement proposals among the learners, the future learning effect of the improvement proposal target person is predicted, and learning is performed according to the prediction result. Make improvement proposals.

Maomi Ueno, “Intelligent LMS with an agent that learns from log data,“ 2005, P.3169-3176

  In the technique shown in Non-Patent Document 1, when a decision tree is used for the above-described model and a plurality of learning results exist as a result under one element, a branch direction with a good learning result among the plurality of results Propose improvement proposals for the contents of elements. For this reason, for example, when the content of the element in the branch direction with a good learning result among a plurality of consequences is the content “to increase the learning time”, an improvement proposal is made with this content.

  Here, the improvement proposal target for which the improvement proposal of “Longer learning time” is made is that the learning time is short, but there is also a person who “dislikes studying” because of the short learning time Can do. It is assumed that a sufficient improvement effect cannot be obtained even if an improvement proposal is made simply to “increase the learning time” for an improvement proposal target person who hates studying.

  As described above, with the technique shown in Non-Patent Document 1, improvement proposals cannot be made with appropriate contents, and the learning effect of the learner may not be improved.

  Therefore, the present invention has been made in view of the above-described problems, and an object thereof is to make an improvement proposal for learning to a learner so that the learning effect can be sufficiently improved.

The present invention proposes the following matters in order to solve the above problems.
(1) The present invention is a learning support device that makes improvement proposals regarding learning to an improvement proposal target person among a plurality of learners, and learning log data that is information when learning is performed, Learning log data collecting means (for example, equivalent to the learning log data collecting unit 10 in FIG. 2) collected for each learner and the learning log data collected by the learning log data collecting means, Learning model construction means (for example, equivalent to the decision tree construction unit 20 in FIG. 2) for constructing a learning model (for example, equivalent to a decision tree described later), a learning model constructed by the learning model construction means, and the learning The difference which calculates | requires the difference of the learning log data of the said improvement proposal object person and a specific results superior group based on the learning log data of the said improvement proposal object person collected by the log data collection means Based on the difference obtained by the obtaining means (for example, the difference obtaining unit 30 in FIG. 2) and the difference obtaining means, the improvement suggesting means (for example, FIG. 2) that makes an improvement proposal to the improvement proposal target person. And the difference acquisition means includes a learning model constructed by the learning model construction means, and a learning log of the improvement proposal target person collected by the learning log data collection means. Learning log data collected by the learning log data collecting means for obtaining a higher grade group composed of persons with better learning results than the improvement proposal target person among the plurality of learners based on the data Based on the above, the improvement proposal target person and the learner belonging to the top grader group are classified into a plurality of groups, and the study is classified into the same group as the improvement proposal target person. The learning support device is characterized in that the specific achievement superior group composed of persons is obtained, and the difference in learning log data between the improvement proposal target person and the specific achievement superior group is obtained. .

  According to the present invention, the learners belonging to the higher grade group composed of learners having better learning results than the improvement proposal target person are classified into the same group as the improvement proposal target person based on the learning log data. And a specific superior group composed of the requested learners. In addition, it was decided to obtain the difference in the learning log data between the improvement proposal target person and the specific achievement superior group. For this reason, it is possible to find a person whose learning status is close to the improvement proposal target person from learners with better learning results than the improvement proposal target person, and to obtain a difference between the found learner and the improvement proposal target person. . Therefore, not only the learning results are better than the improvement proposal target person, but the learning situation with the improvement proposal target person is close (for example, if the improvement proposal target person dislikes studying, the learner dislikes studying) Since the difference from the improvement proposal target person can be obtained, an appropriate improvement proposal can be made to the improvement proposal target person. Therefore, the improvement proposal about learning can be performed with respect to the improvement proposal object person so that a learning effect can fully be improved.

  (2) In the learning support apparatus according to (1), the learning model construction unit constructs a decision tree as the learning model, and the difference acquisition unit (for example, equivalent to the difference acquisition unit 30A in FIG. 6). If the higher grade group includes a plurality of outcomes, the learning log data collection means selects one outcome having the shortest distance from the outcome to which the improvement proposal target person belongs among the plurality of outcomes. Learning based on the learning log data collected by the above, and classifying the improvement proposal target person and a learner belonging to a group composed of learners belonging to the selected outcome into a plurality of groups A support device is proposed.

  Here, based on the improvement proposal from the learning support apparatus, the improvement proposal target person can review the learning method or re-establish the learning plan. However, if the number of improvement proposals becomes too large, there will be improvement proposal candidates who will be less motivated to learn, and improvement proposal candidates who cannot determine which improvement proposal should be dealt with. May not be able to improve the learning effect.

  Therefore, according to the present invention, a decision tree is constructed as a learning model in the learning support apparatus of (1). In addition, when the top grader group includes a plurality of outcomes, one outcome that is closest to the outcome to which the improvement proposal target person belongs is selected from the plurality of outcomes.

  For this reason, it is possible to carefully select an object for which a difference from the improvement proposal target person is to be obtained, so that the number of improvement proposals can be suppressed. Therefore, the learning effect of the improvement proposal target person can be improved.

  (3) In the learning support apparatus according to (1) or (2), the present invention proposes that the improvement proposing means (e.g., corresponding to the improvement proposing part 40A in FIG. 8) is the learning collected by the learning log data collecting means. Log data that is not used for constructing the learning model by the learning model construction unit and whose correlation with the difference obtained by the difference obtaining unit is equal to or greater than a predetermined threshold value Learning based on the difference obtained by the difference obtaining means and information whose correlation with the difference is equal to or greater than the threshold value. A support device is proposed.

  According to the present invention, in the learning support device of (1) or (2), among the collected learning log data and not used for the construction of the learning model, the improvement proposal target person and the specific result superior group It was decided to make a proposal for improvement using the information that was found to be highly correlated with the difference.

  For this reason, it is possible to make multifaceted improvement proposals regarding the differences between the improvement proposal target person and the specific achievement superior group by using information highly correlated with the difference between the improvement proposal candidate and the specific achievement superior group. it can. Therefore, it is possible to facilitate the improvement proposal target person to understand and execute the content of the improvement proposal.

  (4) In the learning support device according to any one of (1) to (3), the learning model construction unit (e.g., corresponding to the decision tree construction unit 20A in FIG. 9) is improved by the improvement proposal unit. A learning support device characterized by determining a value according to an improvement effect on learning of the improvement proposal target person by the proposal being made, and constructing a decision tree having a consequence of the determined number of values is suggesting.

  According to the present invention, in the learning support device according to any one of (1) to (3), the value is determined according to the improvement effect on the learning of the improvement proposal target person as a result of the improvement proposal being made. We decided to build a decision tree with the consequence of the number of values.

  For this reason, the number of outcomes of the decision tree can be changed according to the improvement effect by the improvement proposal. Therefore, since a decision tree according to the improvement effect by the improvement proposal can be constructed, the improvement proposal suitable for the improvement proposal target person can be performed.

  (5) The present invention provides the learning support device according to any one of (1) to (4), wherein the learning log data collection unit uses the learning log data as information based on a learner's physiology during learning (for example, Information on brain waves, pulse, skin conduction, etc., which will be described later, and information on detection results by a 6-axis sensor, which will be described later, and information on the environment around the learner during learning (for example, the temperature around the learner, which will be described later) And a learning support apparatus characterized by using at least one of information on humidity and noise level).

  According to the present invention, in the learning support device according to any one of (1) to (4), as learning log data, information based on a learner's physiology during learning and information about the environment around the learner during learning Or at least one of them.

  For this reason, the situation of the learner who is learning can be grasped in detail and reflected in the improvement proposal. Therefore, the improvement proposal more suitable for the improvement proposal target person can be performed.

  (6) The present invention provides learning log data collection means (for example, equivalent to the learning log data collection unit 10 in FIG. 2), learning model construction means (for example, equivalent to the decision tree construction unit 20 in FIG. 2), difference acquisition means. (For example, equivalent to the difference acquisition unit 30 in FIG. 2) and improvement suggestion means (for example, equivalent to the improvement proposal unit 40 in FIG. 2). A learning support method in a learning support device that makes an improvement proposal, wherein the learning log data collection means collects learning log data, which is information when learning is performed, for each learner; A second step in which the learning model construction means constructs a learning model of the plurality of learners (e.g., corresponding to a decision tree described later) using the learning log data collected in the first step; The difference Based on the learning model constructed in the second step and the learning log data of the improvement proposal target person collected in the first step, the obtaining means and the improvement proposal target person and the specific result high rank A third step of obtaining a difference in learning log data from the person group, and the improvement proposing means making an improvement proposal to the improvement proposal target person based on the difference obtained in the third step. 4, and in the third step, the difference acquisition means learns the learning model constructed in the second step and the improvement proposal target person collected in the first step. Based on the log data, a learning result group collected in the first step is obtained from a plurality of learners who are higher in the learning results than the improvement proposal target person. Based on the data, the improvement proposal target person and the learner belonging to the top grader group are classified into a plurality of groups, and include learners classified into the same group as the improvement proposal target person A learning support method is proposed in which the specific achievement superior group is determined, and a difference in learning log data between the improvement proposal target person and the specific achievement superior group is obtained.

  According to the present invention, the same effects as described above can be obtained.

  (7) The present invention provides learning log data collection means (for example, equivalent to the learning log data collection unit 10 in FIG. 2), learning model construction means (for example, equivalent to the decision tree construction unit 20 in FIG. 2), difference acquisition means. (For example, equivalent to the difference acquisition unit 30 in FIG. 2) and improvement suggestion means (for example, equivalent to the improvement proposal unit 40 in FIG. 2). A program for causing a computer to execute a learning support method in a learning support device that makes an improvement proposal, wherein learning log data that is information when the learning log data collection unit performs learning is stored for each learner. And the learning model construction means uses the learning log data collected in the first step to use the learning models of the plurality of learners (for example, a decision tree described later). The difference acquisition means, the learning model built in the second step, the learning log data of the improvement proposal target collected in the first step, Based on the third step of determining the difference in the learning log data between the improvement proposal target person and the specific achievement superior group, and the improvement suggestion means, based on the difference obtained in the third step, A fourth step of making an improvement proposal to the improvement proposal target person; and in the third step, the difference acquisition means includes a learning model constructed in the second step; Based on the learning log data of the improvement proposal target person collected in the first step, a result composed of persons having a better learning result than the improvement proposal target person among the plurality of learners A superior group is obtained, and based on the learning log data collected in the first step, the improvement proposal target person and a learner belonging to the superior result group are classified into a plurality of groups, and the improvement For obtaining the specific achievement superior group composed of learners classified into the same group as the proposal target person, and for obtaining a difference in learning log data between the improvement proposal candidate and the specific achievement superior group Propose a program.

  According to the present invention, the same effect as described above can be obtained by executing the program using a computer.

  ADVANTAGE OF THE INVENTION According to this invention, the improvement proposal about learning can be performed with respect to a learner so that a learning effect can fully be improved.

1 is a block diagram of a learning support system having a learning support apparatus according to a first embodiment of the present invention. 1 is a block diagram of a learning support apparatus according to a first embodiment of the present invention. It is a figure which shows an example of a decision tree. It is a flowchart which shows operation | movement of the learning assistance apparatus which concerns on 1st Embodiment of this invention. It is a flowchart which shows operation | movement of the learning assistance apparatus which concerns on 1st Embodiment of this invention. It is a block diagram of the learning assistance apparatus which concerns on 2nd Embodiment of this invention. It is a figure which shows an example of a decision tree. It is a block diagram of the learning assistance apparatus which concerns on 3rd Embodiment of this invention. It is a block diagram of the learning assistance apparatus which concerns on 4th Embodiment of this invention. It is a flowchart which shows operation | movement of the learning assistance apparatus which concerns on 4th Embodiment of this invention.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. Note that the constituent elements in the following embodiments can be appropriately replaced with existing constituent elements, and various variations including combinations with other existing constituent elements are possible. Accordingly, the description of the following embodiments does not limit the contents of the invention described in the claims.

<First Embodiment>
[Configuration of Learning Support System AA]
FIG. 1 is a block diagram of a learning support system AA having a learning support apparatus 1 according to the first embodiment of the present invention. The learning support system AA is a system that provides learners with e-learning and online learning services. The learning support device 1, the learning providing device 100, and the learner terminals T1 to Tn (n is an integer satisfying n ≧ 1). ). The learning support device 1, the learning providing device 100, and the learner terminals T1 to Tn are provided so as to be able to communicate with each other via, for example, an Internet line or a telephone line.

  The learning providing apparatus 100 transmits various learning contents to each of the learner terminals T1 to Tn.

  Each of the learner terminals T1 to Tn is provided such that each of the learners U1 to Un can be operated. For example, a PC (abbreviation of Personal Computer), a mobile phone, a tablet PC, or an STB (abbreviation of Set Top Box). ). The learner terminal Tx (x is an arbitrary integer satisfying 1 ≦ x ≦ n) accepts various operations from the learner Ux, outputs video and audio of learning content received from the learning providing device 100, For example, video and audio about the improvement proposal made by the learning support device 1 are output.

  The learning support apparatus 1 makes an improvement proposal for learning, such as a learning method, for an improvement proposal target person among learners who have learned using the learning support system AA. The learning support device 1 will be described in detail below.

[Configuration of Learning Support Device 1]
FIG. 2 is a block diagram of the learning support apparatus 1. The learning support apparatus 1 includes a learning log data collection unit 10, a decision tree construction unit 20, a difference acquisition unit 30, and an improvement proposal unit 40.

  The learning log data collection unit 10 collects learning log data and learning result data for each learner who has learned using the learning support system AA.

  The learning log data is information when learning is performed for each learner. For this learning log data, for example, information such as the time, time zone, and number of times of learning using the learning support system AA, information on the subjects that the learner is good at and not good at are used. It is assumed that the learning log data is acquired by the learning providing device 100 and stored in the learning providing device 100 when the learner performs learning using the learning support system AA.

  The learning result data is information about the learning result in the test taken by the learner, and the test is a test received by the learner using the learning support system AA, a written test, or the like. It is assumed that the learning result data is stored in the learning providing apparatus 100.

  The decision tree construction unit 20 performs learning for each learning content that can be provided by the learning providing device 100 (for example, for each unit such as addition or subtraction) or for each type of learning content (for example, for each subject such as mathematics or English). The decision tree is constructed using the learning log data of all the learners.

FIG. 3 is a diagram illustrating an example of a decision tree constructed by the decision tree construction unit 20. In FIG. 3, i ₁ , i ₂ , i ₂ ′, and i ₃ represent elements, and θ ^ ₁ , θ ^ ₂ , θ ^ ₃ , θ ^ ₄ , and θ ^ ₅ represent consequences. A transition flow that transitions from the element i ₁ to any _one of the consequences θ ^ ₁ to θ ^ ₅ is called a path, and an arrow between elements constituting the path is called a branch.

For example, it is assumed that information indicating whether or not the learning time is 1 hour or more every day among the learning log data collected by the learning log data collecting unit 10 is applied to the element i ₁ . In this case, when the learning time is 1 hour or more every day, “u _i1 = 1”, and when the learning time is less than 1 hour every day, “u _i1 = 0”. Further, for example, it is assumed that information indicating whether or not the time period during which learning was performed is applied to the element i ₂ in the learning log data collected by the learning log data collection unit 10. In this case, if the learning time zone is in the morning, “u _i2 = 1”, and if the learning time zone is in the afternoon, “u _i2 = 0”.

  The difference acquisition unit 30 obtains a difference between the improvement proposal target person and the specific achievement superior group in the decision tree constructed by the decision tree construction unit 20. The specific achievement superior group will be described in detail later.

  The improvement proposing unit 40 transmits the difference obtained by the difference acquiring unit 30 to the learner terminal of the improvement proposing subject and outputs it as video or audio.

[Operation of Learning Support Device 1]
FIG. 4 is a flowchart showing the operation of the learning support apparatus 1.

  In step S1, the learning support apparatus 1 collects learning log data and learning result data by the learning log data collection unit 10, and moves the process to step S2.

  For example, it is assumed that the learning support apparatus 1 makes an improvement proposal for learning using the learning content C1 to the improvement proposal target person. In this case, in step S1, the learning log data when learning is performed with the learning content C1 and the learning result data in the test related to the learning content C1, the improvement proposal target person, all the learners other than the improvement proposal target person, Collect for each of the.

  In step S2, the learning support device 1 performs a decision tree construction process by the decision tree construction unit 20 to construct a decision tree, and moves the process to step S3. Details of the decision tree construction process will be described later with reference to FIG.

  FIG. 5 is a flowchart of the decision tree construction process described above.

  In step S11, the learning support apparatus 1 uses the decision tree construction unit 20 to perform cluster analysis on all the learning result data collected in step S1, classify these learning result data into a plurality, and move the process to step S12. For the cluster analysis, for example, the k-means method can be used.

  In step S12, the learning support apparatus 1 uses the decision tree construction unit 20 to determine each of the classification results in step S11 as a result of the decision tree, and moves the process to step S13.

For example, in FIG. 3, there are five consequences of the consequences θ ^ ₁ to θ ^ ₅ . For this reason, all the learning result data collected in step S1 are classified into five in step S11. In the following, in FIG. 3, learners whose learning results are 0 to 20 are classified as a result θ ^ ₁ , and learners whose learning results are 21 to 40 are classified as a result θ ^ _2. Learners with a learning score of 41 to 60 are classified as an outcome θ ^ ₃ , learners with a learning score of 61 to 80 are classified as an outcome θ ^ ₄ , and a learning score of 81 to 100 It is assumed that the learners up to the point are classified into the result θ ^ ₅ .

  In step S13, the learning support apparatus 1 constructs a decision tree by the decision tree construction unit 20 based on all the learning log data collected in step S1 and the result of the decision tree determined in step S12. The process moves to step S3 in FIG. For example, an algorithm such as CART or C5.0 can be used to construct the decision tree.

Returning to FIG. 4, in step S <b> 3, the learning support device 1 predicts the learning result of the improvement proposal target person using the decision tree constructed in step S <b> 2, and moves the process to step S <b> 4. Specifically, the learning log data of the improvement proposal target person is determined by determining whether the learning log data of the improvement proposal target person collected in step S1 satisfies the contents of each element of the decision tree constructed in step S2. The path on the decision tree according to is identified. According to this, it is estimated from the learning situation of the improvement proposal target person to which of the five outcomes from the result θ ^ ₁ to θ ^ ₅ in FIG. become.

For example, information on whether or not the learning time is 1 hour or more every day is applied to the element i ₁ , and information on whether or not the learning time zone is in the morning is applied to the element i _2. It shall be. Further, it is assumed that the learning log data of the improvement proposal target person includes information that the learning time is less than 1 hour every day and information that the time of learning is in the morning. . Then, the process of step S3 identifies the path that transitions from element i ₁ through element i ₂ to the result θ ^ ₂ as the path on the decision tree corresponding to the learning log data of the improvement proposal target person, and the improvement proposal The subject is estimated to belong to the result θ ^ ₂ . That is, according to the learning situation of the improvement proposal target person so far, when the improvement proposal target person takes the test, it is estimated that the learning result will be between 21 points and 40 points.

  In step S4, the learning support apparatus 1 uses the difference acquisition unit 30 to extract a group composed of learners with better learning results than the improvement proposal target person (hereinafter referred to as “higher grade group”). The processing is moved to S5.

By the process of step S4, in FIG. 3, a group composed of learners belonging to the result θ ^ ₄ and learners belonging to the result θ ^ ₅ is extracted.

  In step S5, the learning support device 1 classifies the improvement proposal target person and the learner belonging to the higher grade group extracted in step S4 into a plurality of groups based on the learning log data collected in step S1. Then, the process proceeds to step S6.

  Specifically, first, among the learning log data collected in step S1, the improvement proposal target person and the learners belonging to the higher grade group extracted in step S4 are acquired. Next, based on each information included in the acquired learning log data, the improvement proposal target person and the learner belonging to the higher grade group extracted in step S4 are classified into a plurality of groups. For the classification, for example, cluster analysis can be used.

  Here, the acquired learning log data includes information that was not used for decision tree construction by the decision tree construction unit 20. For example, it is assumed that information such as whether or not the subject is a good subject is included as information that has been acquired in the above-described learning log data and has not been used to construct a decision tree. Then, if these information is used to classify the above-mentioned improvement proposal target person and learner, the learner who has better learning results than the improvement proposal target person and who is in a learning situation similar to the improvement proposal target person, Can be found.

  In step S6, the learning support device 1 uses the difference acquisition unit 30 to obtain the outcome to which the learner classified into the same group as the improvement proposal target person in step S5 belongs, and a group composed of learners belonging to the obtained outcome. Is obtained as the above-mentioned specific achievement superior group, and the process proceeds to step S7.

  Here, there may be a case where there are a plurality of learners classified in the same group as the improvement proposal target person in step S5, and each of the plurality of learners belongs to different outcomes.

For example, it is assumed that there are four learners classified into the same group as the improvement proposal target person in step S5, three of which belong to the result θ ^ ₄ , and the remaining one belongs to the result θ ^ _5. . In this case, a group composed of learners belonging to the result θ ^ ₄ and learners belonging to the result θ ^ ₅ may be obtained as the above-mentioned specific achievement superior group. In this case, among the learners classified in the same group as the improvement proposal target person in step S5, the result is the result of the largest number of persons belonging to the same outcome, that is, the group composed of learners belonging to the outcome θ ^ _4. May be obtained as the above-mentioned specific achievement superior group.

Further, for example, there are four learners classified into the same group as the improvement proposal target person in step S5, three of them belong to the result θ ^ ₄ , and the remaining one belongs to the result θ ^ ₅ And In addition, learners of five in the outcome θ ^ ₄ belong, the consequences θ ^ ₅ it is assumed that the two persons of the learner belongs. In this case, among learners belonging to the result θ ^ ₄ , the ratio of learners classified into the same group as the improvement proposal target person in step S5 is 60%, and among learners belonging to the result θ ^ ₅ , The ratio of learners classified into the same group as the improvement proposal target person in step S5 is 50%. Therefore, the group composed of learners belonging to the result θ ^ ₄ , which has the highest ratio of learners classified in the same group as the improvement proposal target person in step S5, is obtained as the above-mentioned specific achievement superior group. May be.

  In step S7, the learning support apparatus 1 uses the difference acquisition unit 30 to obtain the difference between the improvement proposal target person and the specific achievement superior group obtained in step S6 in the decision tree constructed in step S2, and step S8. Move processing to.

For example, it is assumed that the improvement proposal target person belongs to the result θ ^ ₂ , and a group composed of learners belonging to the result θ ^ ₄ is extracted as the specific achievement superior group in step S6. Then, by the process of step S7, information “u _i1 = 1”, information “u _i2 = 1”, and information “u _i3 = 0” are obtained as the above-described differences. .

Further, for example, the improvement proposal target person belongs to the result θ ^ ₂ , and is composed of a learner belonging to the result θ ^ ₄ and a learner belonging to the result θ ^ ₅ as the specific achievement superior group in step S6. Assume that a group has been extracted. Then, the information of “u _i1 = 1” and the information of “u _i2 ′ = 1” are obtained as the above-described difference by the process of step S7.

  In step S8, the learning support device 1 causes the improvement proposing unit 40 to transmit the difference obtained in step S7 to the learner terminal of the improvement proposing target person and output it as video or audio, and ends the processing shown in FIG. To do.

  According to the learning support apparatus 1 described above, the following effects can be achieved.

  The learning support device 1 finds out and finds a person who has a learning situation close to the improvement proposal target person from among learners belonging to a group of superior students composed of learners who have better learning results than the improvement proposal target person. Find differences between learners and candidates for improvement. For this reason, it is possible not only to have better learning results than those targeted for improvement proposals, but also to find differences from those targeted for improvement proposals for those whose learning status is close to those targeted for improvement proposals. This can be done for people who are candidates for improvement. Therefore, the improvement proposal about learning can be performed with respect to the improvement proposal subject so that a learning effect can fully be improved.

Second Embodiment
[Configuration of Learning Support Device 1A]
FIG. 6 is a block diagram of a learning support apparatus 1A according to the second embodiment of the present invention. The learning support device 1A can be used instead of the learning support device 1 in the learning support system AA shown in FIG. The learning support device 1 </ b> A differs from the learning support device 1 in that a difference acquisition unit 30 </ b> A is provided instead of the difference acquisition unit 30. In the learning support device 1A, the same components as those of the learning support device 1 are denoted by the same reference numerals, and the description thereof is omitted.

  Similar to the difference acquisition unit 30, the difference acquisition unit 30 </ b> A calculates a higher grade group after obtaining a higher grade group, and determines the difference between the improvement proposal target person and the higher grade group by the decision tree construction unit 20. Find in the decision tree constructed by However, the difference obtaining unit 30A selects and selects one of the plurality of outcomes having the shortest distance from the outcome to which the improvement proposal target person belongs, when the higher grade group includes a plurality of outcomes. A group consisting of learners belonging to the outcome is extracted.

For example, the decision tree shown in FIG. 7 is constructed by the decision tree construction unit 20, the improvement proposal target person belongs to the result θ ^ _12, and a group composed of learners with better learning results than the improvement proposal target person is the result. It is assumed that θ ^ ₁₃ , result θ ^ ₁₄ , and result θ ^ ₁₅ are included. In this case, the number of branches connecting the result θ ^ ₁₂ and the result θ ^ ₁₃ is four, the number of branches connecting the result θ ^ ₁₂ and the result θ ^ ₁₄ is five, and the result θ ^ The number of branches connecting ₁₂ and the consequent θ ^ ₁₅ is 5. Therefore, the difference acquisition unit 30A selects the result θ ^ ₁₃ having the smallest number of branches connected to the result θ ^ _12, and extracts a group composed of learners belonging to the result θ ^ ₁₃ as a higher grade group. To do.

  According to the learning support device 1A described above, in addition to the above-described effects that the learning support device 1 can exhibit, the following effects can be achieved.

  Since the learning support device 1A can carefully select a target for which a difference from the improvement proposal target person is desired, the number of improvement proposals can be suppressed. For this reason, the number of improvement proposals is too large, so it is possible to prevent the desire to learn from declining and it becomes impossible to determine which improvement proposal should be dealt with. The learning effect of the target person can be improved.

<Third Embodiment>
[Configuration of Learning Support Device 1B]
FIG. 8 is a block diagram of a learning support apparatus 1B according to the third embodiment of the present invention. The learning support device 1B can be used in place of the learning support device 1 in the learning support system AA shown in FIG. The learning support device 1B is different from the learning support device 1 in that an improvement proposal unit 40A is provided instead of the improvement proposal unit 40. In the learning support device 1B, the same components as those of the learning support device 1 are denoted by the same reference numerals, and description thereof is omitted.

  Similar to the improvement proposal unit 40, the improvement proposal unit 40A transmits the difference obtained by the difference acquisition unit 30 to the learner terminal of the improvement proposal target person and outputs it as video or audio. However, the improvement proposing unit 40A transmits not only the above-described difference but also information highly correlated with the above-described difference to the learner terminal of the improvement proposal target person, and outputs the information as video or audio.

  For example, in the learning log data of the improvement proposal target person collected by the learning log data collecting unit 10, information that the learning time is less than 1 hour every day and the previous login time to the learning support system AA Information that the login interval until the login time is 24 hours or more is included. Further, the information about whether or not the learning time is 1 hour or more every day is included in the difference obtained by the difference acquisition unit 30, and whether or not the login interval to the learning support system AA is less than 24 hours. It is assumed that the information about k is not used for decision tree construction by the decision tree construction unit 20.

  Since the daily learning time is assumed to be longer as the login interval to the learning support system AA is shorter, the information about whether the login interval to the learning support system AA is less than 24 hours is the learning time. Is highly correlated with information about whether or not is 1 hour or more every day. Therefore, the improvement proposing unit 40A collects the learning log data collecting unit 10 as to whether or not the learning interval belonging to the specific achievement superior group is whether or not the login interval to the learning support system AA is less than 24 hours. The obtained learning log data is obtained, and the obtained information is also transmitted to the learner terminal of the improvement proposal target person.

  The correlation described above is obtained by correlation analysis. Specifically, first, the difference obtained by the difference acquisition unit 30 and the learning log data of the improvement proposal target person collected by the learning log data collection unit 10 and not used for the construction of the decision tree A correlation analysis is performed for each of the above. Next, if the correlation analysis result is greater than or equal to a predetermined threshold, it is determined that the correlation with the difference obtained by the difference acquisition unit 30 is high.

  In addition, when there are a plurality of items determined to be highly correlated with the difference obtained by the difference acquisition unit 30, all of them may be transmitted to the learner terminal of the improvement proposal target person, Only those having high correlation may be transmitted to the learner terminal of the improvement proposal target person.

  According to the learning support device 1B described above, in addition to the above-described effects that the learning support device 1 can exhibit, the following effects can be achieved.

  The learning support apparatus 1B multifacetedly proposes improvement proposals regarding differences between the improvement proposal target person and the specific achievement superior group, based on information highly correlated with the difference between the improvement proposal candidate and the specific achievement superior group. It can be carried out. Therefore, it is possible to facilitate the improvement proposal target person to understand and execute the content of the improvement proposal.

<Fourth embodiment>
[Configuration of Learning Support Device 1C]
FIG. 9 is a block diagram of a learning support apparatus 1C according to the fourth embodiment of the present invention. The learning support device 1C can be used in place of the learning support device 1 in the learning support system AA shown in FIG. The learning support device 1 </ b> C differs from the learning support device 1 in that it includes a decision tree construction unit 20 </ b> A instead of the decision tree construction unit 20. In the learning support device 1C, the same constituent elements as those of the learning support device 1 are denoted by the same reference numerals, and the description thereof is omitted.

  Similar to the decision tree construction unit 20, the decision tree construction unit 20A constructs a decision tree using the learning log data. However, the decision tree construction unit 20A changes the number of consequences of the decision tree according to the improvement effect by the improvement proposal.

  FIG. 10 is a flowchart of the decision tree construction process according to this embodiment.

  In step S21, the learning support device 1C determines whether or not a sufficient improvement effect has not been obtained by the previous improvement proposal by the decision tree construction unit 20A. If it is determined that it has not been obtained, the process proceeds to step S22. If it is determined that it has been obtained, or if the current improvement proposal is the first time, the process proceeds to step S23.

  Here, the determination whether or not a sufficient improvement effect has been obtained is performed using the learning log data of the improvement proposal target person collected by the learning log data collection unit 10. Specifically, among the learning log data of the improvement proposal target person collected by the learning log data collection unit 10, information indicating the improvement proposal transmitted to the improvement proposal target person's learner terminal in the previous improvement proposal in advance Consider whether or not improvement has been made beyond a defined threshold. Then, when the improvement over the threshold is performed, it is determined that a sufficient improvement effect is obtained, and when the improvement over the threshold is not performed, it is determined that the sufficient improvement effect is not obtained.

  In step S22, the learning support device 1C causes the decision tree construction unit 20A to increase the number k resulting from the decision tree by n (k is an integer that satisfies k ≧ 2 and n is an integer that satisfies n ≧ 1). ), The process proceeds to step S23.

  In step S23, the learning support device 1C causes the decision tree construction unit 20A to perform all learning result data collected in step S1 in the same manner as the processing performed by the learning support device 1 in the decision tree construction unit 20 in step S11 of FIG. On the other hand, these learning result data are classified into a plurality, and the process proceeds to step S24. The value of “k” described above is applied to the number for classifying the learning result data.

  In each of step S24 and step S25, the learning support apparatus 1C performs the same process as the process performed by the learning support apparatus 1 in each of step S12 and step S13 in FIG.

  According to the learning support device 1C described above, in addition to the above-described effects that the learning support device 1 can exhibit, the following effects can be achieved.

  The learning support apparatus 1C can change the number of outcomes of the decision tree according to the improvement effect by the improvement proposal. Therefore, since a decision tree according to the improvement effect by the improvement proposal can be constructed, the improvement proposal suitable for the improvement proposal target person can be performed.

  Note that the processing of the learning support devices 1, 1A, 1B, and 1C of the present invention is recorded on a computer-readable non-transitory recording medium, and the program recorded on the recording medium is stored in the learning support devices 1, 1A, 1B. 1C can be read and executed, and the present invention can be realized.

  Here, for example, a nonvolatile memory such as an EPROM or a flash memory, a magnetic disk such as a hard disk, a CD-ROM, or the like can be applied to the above-described recording medium. Further, reading and execution of the program recorded on the recording medium is performed by a processor provided in the learning support devices 1, 1A, 1B, and 1C.

  The above-described program is transmitted from the learning support apparatuses 1, 1A, 1B, and 1C storing the program in a storage device or the like to another computer system via a transmission medium or by a transmission wave in the transmission medium. May be. Here, the “transmission medium” for transmitting the program refers to a medium having a function of transmitting information, such as a network (communication network) such as the Internet or a communication line (communication line) such as a telephone line.

  Further, the above-described program may be for realizing a part of the above-described function. Furthermore, what can implement | achieve the above-mentioned function in combination with the program already recorded on the learning assistance apparatus 1, 1A, 1B, 1C, what is called a difference file (difference program) may be sufficient.

  The embodiment of the present invention has been described in detail with reference to the drawings. However, the specific configuration is not limited to this embodiment, and includes a design that does not depart from the gist of the present invention.

  For example, in the first embodiment described above, the decision tree is constructed every time the improvement proposal target person is made the improvement proposal as shown in FIG. It may be performed at a predetermined timing, such as once a day, or each time an improvement proposal is made a predetermined number of times.

  In the first embodiment described above, the learning result of the improvement proposal target person is predicted using the decision tree. However, the present invention is not limited to this. For example, a test result received by the improvement proposal target person may be used as the learning result of the improvement proposal target person.

  Further, in the above-described third embodiment, the difference acquisition unit 30 among the information that is the improvement log target learning log data collected by the learning log data collection unit 10 and was not used for the construction of the decision tree. Those having a high correlation with the obtained difference were determined by correlation analysis. Here, this correlation analysis may be performed every time a difference is obtained by the difference acquisition unit 30, or is performed in advance, and it is determined that which information in the learning log data is highly correlated with which information. It may be a thing.

  In the fourth embodiment described above, the number k of decision tree outcomes is increased by n in step S22. However, the present invention is not limited to this, and the number k of outcomes of the decision tree is decreased by n. Also good.

  In each embodiment described above, the learning environment data may also be used as part of the learning log data. Here, the learning environment data is information based on the physiology such as the learner's brain wave, pulse, and skin conduction during learning, information on the ambient temperature, humidity, and noise level of the learner during learning, It is the information of the detection result by the 6-axis sensor of the learner in the inside and the learner terminal, and is acquired by various sensors provided in the learner terminal. The learning environment data is also used as a part of the learning log data, so that the learning situation of the learner can be grasped more finely and an improvement proposal can be made.

AA ... Learning support system 1, 1A, 1B, 1C ... Learning support device 10 ... Learning log data collection unit 20, 20A ... Decision tree construction unit 30, 30A ... Difference acquisition unit 40, 40A ... Improvement Proposal Department

Claims (7)

A learning support device that makes improvement proposals for learning to a candidate for improvement proposal among a plurality of learners,
Learning log data collecting means for collecting learning log data, which is information at the time of learning, for each learner;
Learning model construction means for constructing learning models of the plurality of learners using the learning log data collected by the learning log data collection means;
Based on the learning model constructed by the learning model construction means and the learning log data of the improvement proposal target person collected by the learning log data collection means, the improvement proposal target person and the specific result superior group A difference acquisition means for obtaining a difference in learning log data of
Improvement proposal means for making an improvement proposal for the improvement proposal target person based on the difference obtained by the difference acquisition means, and
The difference acquisition means includes
Based on the learning model constructed by the learning model construction means and the learning log data of the improvement proposal target person collected by the learning log data collection means, the improvement proposal target person among the plurality of learners Seeking a higher grade group consisting of those with better learning results than
Based on the learning log data collected by the learning log data collection means, classify the improvement proposal target person and learners belonging to the higher grade group into a plurality of groups,
Seeking the specific achievement superior group composed of learners classified into the same group as the improvement proposal target person,
A learning support apparatus characterized by obtaining a difference in learning log data between the improvement proposal target person and the specific achievement superior group. The learning model construction means constructs a decision tree as the learning model,
The difference acquisition means, when the higher grade group includes a plurality of consequences,
Select one of the plurality of outcomes that has the shortest distance from the outcome to which the improvement proposal target person belongs,
Based on the learning log data collected by the learning log data collecting means, the improvement proposal target person and learners belonging to a group composed of learners belonging to the selected outcome are classified into a plurality of groups. The learning support apparatus according to claim 1. The improvement suggestion means is:
Correlation between the learning log data collected by the learning log data collecting means and the difference obtained by the difference obtaining means among the information not used for constructing the learning model by the learning model construction means That is greater than or equal to a predetermined threshold,
The improvement proposal is made to the improvement proposal target person based on the difference obtained by the difference acquisition means and information on which the correlation between the difference is equal to or greater than the threshold value. Or the learning support apparatus according to 2. The learning model construction means includes
In accordance with the improvement effect on learning of the improvement proposal target person by the improvement proposal by the improvement proposal means, a value is determined,
The learning support apparatus according to claim 1, wherein a decision tree having a consequence of the determined number of values is constructed.   The learning log data collection means uses, as the learning log data, at least one of information based on a learner's physiology during learning and information on an environment around the learner during learning. The learning support apparatus according to claim 1. A learning support method in a learning support apparatus comprising learning log data collection means, learning model construction means, difference acquisition means, and improvement suggestion means, and makes improvement proposals for learning among improvement learners among a plurality of learners There,
A first step in which the learning log data collection means collects learning log data, which is information at the time of learning, for each learner;
A second step in which the learning model construction means constructs a learning model of the plurality of learners using the learning log data collected in the first step;
The difference acquisition means identifies the improvement proposal target person based on the learning model constructed in the second step and the learning log data of the improvement proposal target person collected in the first step. A third step of finding a difference in learning log data from the higher grade group;
The improvement proposal means comprises a fourth step of making an improvement proposal to the improvement proposal target person based on the difference obtained in the third step;
In the third step, the difference acquisition means
Based on the learning model constructed in the second step and the learning log data of the improvement proposal target person collected in the first step, from the improvement proposal target person among the plurality of learners Also asked for a higher grade group consisting of those with good learning results,
Based on the learning log data collected in the first step, classify the improvement proposal target person and learners belonging to the higher grade group into a plurality of groups,
Seeking the specific achievement superior group composed of learners classified into the same group as the improvement proposal target person,
A learning support method, comprising: obtaining a difference in learning log data between the improvement proposal target person and the specific achievement superior group. A learning support method in a learning support apparatus comprising a learning log data collection means, a learning model construction means, a difference acquisition means, and an improvement proposal means, and makes improvement proposals for learning to an improvement proposal target person among a plurality of learners A program for causing a computer to execute,
A first step in which the learning log data collection means collects learning log data, which is information at the time of learning, for each learner;
A second step in which the learning model construction means constructs a learning model of the plurality of learners using the learning log data collected in the first step;
The difference acquisition means identifies the improvement proposal target person based on the learning model constructed in the second step and the learning log data of the improvement proposal target person collected in the first step. A third step of finding a difference in learning log data from the higher grade group;
The improvement proposal means causes the computer to execute a fourth step of making an improvement proposal to the improvement proposal target person based on the difference obtained in the third step;
In the third step, the difference acquisition means
Based on the learning model constructed in the second step and the learning log data of the improvement proposal target person collected in the first step, from the improvement proposal target person among the plurality of learners Also asked for a higher grade group consisting of those with good learning results,
Based on the learning log data collected in the first step, classify the improvement proposal target person and learners belonging to the higher grade group into a plurality of groups,
Seeking the specific achievement superior group composed of learners classified into the same group as the improvement proposal target person,
The program for calculating | requiring the difference of the learning log data of the said improvement proposal object person and the said specific results superior group.

Images