KR102126718B1 - System and method of providing custom-made learning contents using artificial intelligence - Google Patents

Abstract

The disclosure disclosed herein relates to a system and method for providing customized learning content using artificial intelligence, and more particularly, to a system and method for providing customized learning content to a user according to a user's memory and difficulty of learning content. It is about.
In more detail, the disclosed contents include a content management unit for storing data related to the learning content, a learning history big data management unit for storing data regarding a user's learning history, and the content management unit in a learning content providing system using artificial intelligence. In the learning optimization engine unit, the learning optimization engine unit calculates and corrects the memory retention amount for each user's content according to the data related to the learning contents stored in the data and the learning history data of the user stored in the learning history big data management unit. Based on the calculated and corrected memory retention amount for each content, a content presentation unit that provides a user with a memory storage amount in descending order, a content scoring unit that scores a user's learning result for the learning content, and the user's learning results Accordingly, the present invention provides a system for providing customized learning content using artificial intelligence, including a content difficulty correcting unit that corrects the difficulty of the learning content.

Description

System and method for providing custom-made learning contents using artificial intelligence {System and method of providing custom-made learning contents using artificial intelligence}

The disclosure disclosed herein relates to a system and method for providing customized learning content using artificial intelligence, and more particularly, to a system and method for providing customized learning content to a user according to a user's memory and difficulty of learning content. It is about.

Unless otherwise indicated herein, the content described in this section is not prior art to the claims of this application and is not admitted to be prior art by inclusion in this section.

In general, the method of providing learning content is mainly to listen to an institute or lecture offline, or to learn through a book, but recently, data processing technology through a communication network such as a high-speed Internet or a mobile communication network has been developed, and the Internet, etc. The provision of lectures and learning content using only communication is increasing. In addition, with the development of smartphones, regardless of the location, it is possible to provide a variety of learning content.

Accordingly, lectures conducted offline can also be taken online, and as online learning becomes possible in real time, the number of websites providing such learning content has increased. In particular, as the market for private education expands, and social learning efforts to acquire various professional qualifications and linguistic qualifications, online learning contents regardless of time and place are developing in a form that reflects the needs of consumers.

However, the existing method of providing learning content is limited to a method of playing a video lecture online or displaying the document by documenting the lecture content. Further, even if the user provides content to solve the problem online, it simply provides the problem data stored on the database and does not take into account the user's characteristics at all.

In addition, the difficulty of the provided problem is arbitrarily classified, and it is fixed uniformly, so there is a problem that does not reflect the actual user's perceived difficulty, and the problem of the incorrectly set difficulty causes a distorted learning result.

In Korean Patent Publication No. 10-2018-0049468, "E-learning learning content providing system and method", as an e-learning learning content providing method for step-by-step learning and evaluation to users through an online environment, it provides learning content selected by the user and , Provides a problem assigned to the corresponding content, and accordingly, calculates a correct answer rate to determine whether the user has corrected the problem and provides customized content to the user. However, in order to provide customized content to the user only with the correct answer rate, the complexity of the user is overlooked, and there is a need for a more accurate and systematic method of providing customized content.

1. Korea Patent Publication No. 10-2018-0049468

By using AI (artificial intelligence) and big data, we assign customized oblivion curves to each user, and we want to provide customized learning content to provide the best learning effect to users.

Disclosed is a learning content providing system using artificial intelligence, the content management unit storing data related to the learning content, the learning history big data management unit storing data related to the user's learning history, the learning content stored in the content management unit A learning optimization engine unit that calculates and corrects a user's memory retention amount for each content according to data about the user and learning history of the user stored in the learning history big data management unit, and calculated and corrected by the learning optimization engine unit Based on the amount of memory reserved for each content, a content presentation unit that provides users with a low memory storage amount, a content scoring unit that scores the user's learning results for the learned learning content, and the learning content according to the user's learning results. The content difficulty correcting unit for correcting the difficulty level is presented as an embodiment of a user-customized learning content providing system using artificial intelligence.

In addition, as another embodiment, in a method for providing user-customized learning content using artificial intelligence, a first step of receiving a learning content call from a user's terminal, a second step of extracting a user's memory retention amount for each content, the user The third step of correcting the memory retention amount of each content according to the user's learning history data, the fourth step of extracting and arranging the basic learning content in the order of the user's memory retention amount, the user's application content submission rate The fifth step of extracting, the sixth step of correcting the user's application content submission rate according to the data of the user's learning history, the user's application content submission rate of the basic learning content arranged in the order of the lowest storage amount for each content According to the present invention, a seventh step of converting to applied learning content and an eighth step of providing the arranged basic learning content and applied learning content to a user's terminal are provided.

According to the disclosed embodiment, it is possible to provide customized learning content that is most suitable for the user by analyzing the big data related to the user's learning history by artificial intelligence, and in particular, presenting a personalized oblivion curve to analyze the user's memory retention amount By doing so, it is possible to present learning content for a long memory.

The effects of the present embodiments are not limited to the effects mentioned above, and other effects not mentioned will be clearly understood by those skilled in the art from the description of the claims.

1(a) is an explanatory diagram of the oblivion curve of the ebbing house.
Figure 1 (b) is an exemplary view of the user's customized oblivion curve according to an embodiment of the disclosed content.
2 is a schematic diagram of a system for providing customized learning content according to an embodiment of the disclosed content.
3 is a schematic diagram of a learning optimization engine unit according to an embodiment of the disclosed subject matter.
Figure 4 is a graph and a table for an embodiment of correcting the amount of storage for each content.
5 is a flowchart of a method for providing user-customized learning content according to an embodiment of the disclosed content.
Figure 6 (a) is a flow chart of a user's content storage capacity correction according to an embodiment of the disclosed content.
Figure 6 (b) is a flow diagram of a user's application content submission rate correction according to an embodiment of the disclosed content.
7 is a flow chart of a terminal and a server according to an embodiment of the disclosed subject matter.

Advantages and features of the present invention, and methods for achieving them will be clarified with reference to embodiments described below in detail together with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but may be implemented in various different forms, and only the embodiments allow the disclosure of the present invention to be complete, and common knowledge in the technical field to which the present invention pertains. It is provided to fully inform the holder of the scope of the invention, and the invention is only defined by the scope of the claims. Throughout the specification, the same reference numerals refer to the same components.

In describing embodiments of the present invention, when it is determined that a detailed description of known functions or configurations may unnecessarily obscure the subject matter of the present invention, the detailed description will be omitted. In addition, the terms described below are terms defined in consideration of functions in an embodiment of the present invention, which may vary according to a user's or operator's intention or practice. Therefore, the definition should be made based on the contents throughout this specification.

Prior to describing the embodiments, the contents for helping to understand the contents posted in the present specification will be first described.

The goal of users who use learning content is to bring learning content to long-term memory. However, when learning, users forget their learning contents, and it is the Ebbing House oblivion curve that is schematized. However, each user's unique memory ability is different from each other, and when providing learning content, a method of providing customized content is required according to the user's forgetting rate, that is, the amount of memory retained.

As shown in Fig. 1(a), the user first learns the memory to show a constant graph over time, which is a general oblivion curve. Looking at the graph, forgetting starts 10 minutes after the user learns, and after 1 day, 70% or more of the learning content is forgotten, and after 1 month, 80% of the learning content is forgotten. Therefore, an effective learning method is a method of newly learning before oblivion occurs. Looking at the graph (B) of FIG. 1(a), if the user repeatedly learns before much forgetting progresses, the learning contents can be memorized for a long time.

The oblivion curve is summarized by the following formula.

In the forgetting curve formula, R means the memory storage amount for each general content, t is time, and S is the general memory coefficient. S is a generalized memory coefficient, and the general average user's general memory coefficient can be quantified to about 71.52156.

Hereinafter, the oblivion curve of the above-described ebbing house may be processed to be used in an embodiment of the present invention, and the variables used in the embodiment of the present invention will be defined.

In the above forgetting curve formula, if the unit of time is adjusted to one day,

It can be defined as, D in the above formula means a day (1day) unit.

The general memory coefficient may be tailored to the user according to the memory power correction coefficient according to the user's learning ability.

More specifically, the user's corrected memory coefficient is

는 기억력 보정계수를 의미한다. 상기 기억력 보정계수는 일반 기억력을 기준으로 문제의 정답비율을 추정하고, 상기 추정된 정답비율에 의해 추정된 기대정답개수와 실제 정답개수의 차이를 기준으로 사용자별 기억력을 보정할 수 있는 계수이며, 아래와 같이Can be defined as

Means the memory correction factor. The memory correction coefficient is a coefficient capable of estimating a correct answer ratio of a problem based on a general memory, and correcting memory for each user based on a difference between an expected number of correct answers estimated by the estimated correct answer rate and an actual number of correct answers, As follows

는 기억력 오차를 의미하며, 상기 기억력 오차는 Can be defined as remind

Means memory error, and the memory error is

는 정답개수를 의미하고, 상기

는 기대정답개수를 의미한다. 상기 정답개수는 사용자에게 제공된 문제개수를 사용자가 풀고난 후에 채점을 통한 정답개수를 의미하며, 상기 기본 학습 콘텐츠와 상기 응용 학습 콘텐츠를 포함한다.Can be defined as remind

Means the number of correct answers, and

Means the number of expected correct answers. The number of correct answers means the number of correct answers through scoring after the user solves the number of questions provided to the user, and includes the basic learning content and the applied learning content.

The expected number of correct answers is

은 문제개수를 의미한다. 상기 문제개수는 본 발명의 일 실시 예에 따라 제공한, 상기 기본 학습 콘텐츠와 상기 응용 학습 콘텐츠의 총 합을 의미한다.Can be defined as

Means the number of problems. The number of problems means a total sum of the basic learning content and the applied learning content provided according to an embodiment of the present invention.

Therefore, by substituting the user's learning result into the expected number of correct answers, the memory error and the memory correction factor, the general memory coefficient can be corrected by the user's corrected memory coefficient, and the user's corrected memory coefficient In line with this, it is possible to provide a customized oblivion curve based on the user's memory.

The memory correction coefficient of the user may be measured when learning is started, and may be periodically measured when data related to the user's learning history is accumulated.

Hereinafter, a preferred embodiment of a system for providing user-specific learning content using artificial intelligence will be described with reference to the above-mentioned contents and the accompanying drawings.

2 and 3, the user-customized learning content providing system using artificial intelligence includes a content management unit 10, a learning history big data management unit 20, a learning optimization engine unit 30, and a content presentation unit 40 ), a content scoring unit 50 and a content difficulty correction unit 60.

The content management unit 10 may store data related to learning content. The data related to the learning content may include the basic learning content 11, the applied learning content 12, and the difficulty level 13 for each content, and may also include any information regarding learning content available to other users. . The difficulty level 13 for each content can be divided into three or more grades, and as an example, can be divided into grades 1 to 5, and the difficulty of the problem is quantified, and the grade 1 is within the top 8.8%. , 2nd grade can be divided into the top 20.2%, 3rd grade is the top 39.2%, 4th grade is the top 65.8%, and 5th grade is the top 100%.

The learning history big data management unit 20 may store data related to a user's learning history. The data on the user's learning history includes the user's oblivion curve (21), learning date (22), learning achievement rate (23), memory retention by content (24), application content submission rate (25), and learning result (26). It may include, and is not limited to the above configurations, and may include all data related to a user's learning history.

The user's oblivion curve 21 stored in the learning history big data management unit 30,

By the amount of memory reserved for each general content expressed by R, the time expressed by t, and the general memory coefficient, which is expressed by S and can be defined as 71.52156,

In the general oblivion curve defined as, reflecting the user's learning results (26),

Can be defined as The R'refers to the user's content storage amount 24, and the S'refers to the corrected memory coefficient. The corrected memory coefficient is

는 정답개수를 의미하고,

은 문제개수를 의미한다. 상기 정답개수는 사용자에게 제공된 문제개수를 사용자가 풀고난 후에 채점을 통한 정답개수를 의미하며, 상기 기본 학습 콘텐츠와 상기 응용 학습 콘텐츠를 포함하며, 상기 문제개수는 본 발명의 일 실시 예에 따라 제공한, 상기 기본 학습 콘텐츠와 상기 응용 학습 콘텐츠의 총 합을 의미한다.Can be defined as

Means the number of correct answers,

Means the number of problems. The number of correct answers means the number of correct answers through scoring after the user solves the number of questions provided to the user, and includes the basic learning content and the applied learning content, and the number of questions is provided according to an embodiment of the present invention In addition, it means the total sum of the basic learning content and the applied learning content.

로 표현하는 상기 기억력 보정계수로 표현가능하며, Where the corrected memory coefficient is

It can be expressed by the memory correction coefficient represented by

는 기억력 오차를 의미하며, 상기 기억력 오차는 Can be defined as

Means memory error, and the memory error is

는 기대정답개수를 의미하며,Can be expressed as remind

Means the number of correct answers,

Can be defined as The user's application content submission rate (25) is

는 응용 콘텐츠 정답개수를 의미하고,

은 응용 콘텐츠 개수를 의미한다. 상기 응용 콘텐츠 개수는 본 발명의 일 실시 예에 따라 제공한 전체 학습 콘텐츠 중에 상기 응용 학습 콘텐츠의 수를 의미하며, 상기 응용 콘텐츠 정답개수는 제공된 상기 응용 학습 콘텐츠를 사용자가 풀고난 후 채점을 통한 정답개수를 의미한다.It is defined as, the M'refers to the user's application content submission rate 25, the M refers to the general application content submission rate,

Means the number of correct answers for application content,

Means the number of application contents. The number of application contents means the number of the application learning contents among all the learning contents provided according to an embodiment of the present invention, and the number of correct answers to the application contents is correct through scoring after the user solves the provided application learning contents It means the number.

The learning optimization engine unit 30 according to the data related to the learning content stored in the content management unit 10 and the learning history of the user stored in the learning history big data management unit 20, the content of the user It is possible to calculate and correct the memory storage amount 24 for each star.

At this time, the storage amount (R') 24 of each user's content is

Or, adjust the time unit to one day (1Day),

It may be defined as, the user's corrected memory coefficient is

는 정답개수를 의미하고,

은 문제개수를 의미한다. 상기 정답개수는 사용자에게 제공된 문제개수를 사용자가 풀고난 후에 채점을 통한 정답개수를 의미하며, 상기 기본 학습 콘텐츠와 상기 응용 학습 콘텐츠를 포함하며, 상기 문제개수는 본 발명의 일 실시 예에 따라 제공한, 상기 기본 학습 콘텐츠와 상기 응용 학습 콘텐츠의 총 합을 의미한다. 따라서, 상기 학습결과(26)에 의해 보정된 기억력 계수를 보정할 수 있다. 보다 상세하게 살펴보면, 상기 학습이력 빅데이터 관리부(30)에 저장된 상기 학습성취율(23) 및 학습결과(26)와 상기 콘텐츠 관리부(10)에 저장된 상기 콘텐츠별 난이도(13)에 따라, 상기 콘텐츠별 기억보유량(24)을 보정하고, 보정된 상기 콘텐츠별 기억보유량(24)에 따라 각 사용자의 상기 망각곡선(21)을 보정할 수 있다.It can be defined as, as described above,

Means the number of correct answers,

Means the number of problems. The number of correct answers means the number of correct answers through scoring after the user solves the number of questions provided to the user, and includes the basic learning content and the applied learning content, and the number of questions is provided according to an embodiment of the present invention In addition, it means the total sum of the basic learning content and the applied learning content. Therefore, the memory power coefficient corrected by the learning result 26 can be corrected. Looking in more detail, according to the learning achievement rate 23 and learning result 26 stored in the learning history big data management unit 30 and the difficulty of each content (13) stored in the content management unit 10, each content The storage amount 24 may be corrected, and the oblivion curve 21 of each user may be corrected according to the corrected storage amount 24 for each content.

In general, for the first learning content, forgetting starts by the general forgetting curve or the user's forgetting curve, and the user's memory storage amount by content is reduced by the forgetting curve. However, even if the learning content has already been learned by the user and the learning achievement rate 23 of the user for the learning content is high, if the memory retention amount of each user's content is determined by the forgetting curve of the user, it is actually There is a concern that the amount of memory held by each user's content and the sense of separation may occur. For example, if a user learns a specific content once, and then a question about the content is presented once or several times, and the user corrects the answer, the user will be highly evaluated for learning achievement of the specific content. I think you remember a lot. In this case, it would be inappropriate to evaluate the memory retention amount for the corresponding content low by simply learning the corresponding content once.

Accordingly, in order to correct such an error, it is necessary to correct the memory retention amount of each user's content in accordance with the learning achievement rate grade by the learning achievement rate 23 for each content. As a method of correcting, as an embodiment of the present invention, when the learning achievement rate grade is high, that is, when the correct answer rate for the corresponding content is high, the corresponding learning content may be considered as reviewed, and the learning achievement rate grade An arbitrary period of 1 day or more may be determined, and the memory retention amount of each user's content at a certain time point may be progressively corrected according to each learning achievement rate level. Assuming that an average of 10 days is required for the first-stage update of the learning achievement rate level for each content, the period during which the learning achievement rate level for each content is maintained may be replaced with a corresponding learning repetition cycle.

With reference to the graph (a) of FIG. 4, it is as follows. If a user's learning achievement rate for a specific content is first grade 1, the user's memory retention amount for the corresponding content is quantified to about 0.304224, and then, as the daily date passes, the memory retention rate for the content is considered to decrease at a certain rate. If a random cycle in which learning is repeated is set to 10 days (the learning achievement rate of each user's content is estimated to be 1st grade during the learning repeating cycle (10 days)), then the next learning by questioning and correct answer evaluation of the content in the meantime If the learning achievement rate of the user has risen to the second grade in the period, it is determined that the user has significantly remembered the corresponding content, and therefore it is necessary to correct the user's memory retention amount for each content higher.

Therefore, the corrected user's memory retention amount for each content is calculated based on "the first set learning achievement rate (1st grade in this example), and the memory retention amount for each user's content reduced over time" changes from the beginning of the next cycle. It can be expressed by summing the user's memory retention amount for each content calculated based on the learning achievement rate.

The learning achievement rate grade consists of 5 stages, and the learning achievement rate is evaluated and updated every 10 days. For example, if the initial learning achievement rate grade is 1 grade, the time elapsed after learning and the learning achievement rate Depending on the grade change, the user's content storage amount may be corrected as follows.

Memory retention in the first cycle

In the second cycle, if the learning achievement rate level is 2

(Where D represents the date that has elapsed since the learning achievement rate for the content was changed to 2)

In the third cycle, if the learning achievement rate grade is 3

(Where D represents the date that has elapsed since the learning achievement rate for the content was changed to 3)

In the fourth cycle, if the learning achievement rate grade is 4

(Where D represents the date that has elapsed since the learning achievement rate grade for the content was changed to 4 grade)

If the learning achievement rate level has risen to level 5 since then

It is defined as, and it can be classified that the memory retention amount is set to 1 and the user is no longer asked because the complete learning on the corresponding content has been achieved.

Regarding the method of correcting the memory retention amount for each content according to the learning achievement rate level of the user, the generalization thereof is summarized as follows.

The G refers to the learning achievement rate grade for each user's content, the n refers to an integer that increases by 1, and the P refers to an arbitrary cycle of one or more days during which learning is repeated. In the above example, it means 10 days.

In this way, the weight of the memory holding amount is obtained for each individual content, and among the individual contents, content that needs to be first learned can be presented to the user. That is, the learning effect can be improved by inducing the user to review contents having a low memory retention amount.

In addition, under the assumption that the more difficult the learning content is, the more difficult it is to forget, and by multiplying the coefficients corresponding to the difficulty 13 for each content divided into three or more ratings, the memory retention amount 24 for each content can be corrected. .

The content questionnaire 40 may provide the user with a memory storage amount in descending order based on the memory storage amount 24 for each content calculated and corrected by the learning optimization engine unit 30. More specifically, the basic learning content 11 having a low memory retention amount is preferentially extracted based on the user's memory storage amount 24 for each content, and the extracted basic learning content 11 is the user's respective user. In accordance with the application content submission rate of 25, it can be converted into the application learning content 12 and provided, and the application content submission rate 25 can be corrected by the learning result 26.

The application content submission rate 25 represented by M'is

는 응용 콘텐츠 정답개수를 의미하고,

은 응용 콘텐츠 개수를 의미하며,

은 문제개수를 의미한다. 상기 문제개수는 본 발명의 일 실시 예에 따라 제공한, 상기 기본 학습 콘텐츠와 상기 응용 학습 콘텐츠의 총 합을 의미하며, 상기 응용 콘텐츠 개수는 본 발명의 일 실시 예에 따라 제공한 전체 학습 콘텐츠 중에 상기 응용 학습 콘텐츠의 수를 의미하며, 상기 응용 콘텐츠 정답개수는 제공된 상기 응용 학습 콘텐츠를 사용자가 풀고난 후 채점을 통한 정답개수를 의미한다.Can be defined as The M refers to the general application content submission rate,

Means the number of correct answers for application content,

Means the number of application contents,

Means the number of problems. The number of problems means the total sum of the basic learning content and the applied learning content provided according to an embodiment of the present invention, and the number of the applied content is among the total learning content provided according to an embodiment of the present invention. It means the number of the application learning contents, and the number of correct answers of the application contents means the number of correct answers through scoring after the user solves the provided application learning content.

는 기대정답개수를 의미하며,remind

Means the number of correct answers,

Is defined as In this case, when the learning achievement rate 23 of the user's application content submission rate 25 is low, not providing the applied learning content 12 may increase the learning effect.

Therefore, the general application content submission rate varies according to the learning achievement rate 23, and when the learning achievement rate 23 is divided into three or more grades,

It may be defined as, and through this, the user's application content submission rate 25 may be corrected.

The content scoring unit 50 may score a user's learning result 25 for the presented learning content. The learning results 25 scored by the content scoring unit 50 may be stored in the learning history big data management unit 20.

The content difficulty correction unit 60 may correct the difficulty of the learning content according to the learning result 26 of the user.

Hereinafter, a method for providing user-customized learning content using artificial intelligence will be described.

5, 6 (a), as shown in Figure 6 (b), in a method for providing user-specific learning content using artificial intelligence, a first step of receiving a learning content call from a user's terminal (S100), A second step (S200) of extracting the user's memory retention amount for each content (S200), a third step of correcting the user's memory retention amount for each content according to data related to the user's learning history (S300), and the user The fourth step (S400) of extracting and arranging the basic learning contents 11 in the order of the lowest memory storage amount (24) of each content, the fifth step (S500) of extracting the application content submission rate 25 of the user, the user The sixth step (S600) of correcting the application content questionnaire 25 of the user according to the data related to the learning history of the user, and the user having the basic learning content 11 arranged in the order of the lowest storage amount 24 for each content A seventh step (S700) of converting the applied learning content 12 into the applied learning content 12 according to the application content submission rate of 25, and providing the arranged basic learning content 11 and the applied learning content 12 to the user's terminal And receiving learning results 26 for the basic learning content 11 and the applied learning content 12 provided to the user's terminal in step S800 and the eighth step S800, wherein the learning And a ninth step (S900) of correcting the difficulty levels of the basic learning content 11 and the applied learning content 12 according to the result 26.

If you look specifically,

In the first step S100, a learning content call is received from the user's terminal. It includes any configuration that can communicate with a user's terminal, and any method capable of communicating, such as wired communication or wireless communication, can be used in the communication method.

In the second step (S200 ), the user's content storage amount 24 is extracted.

In the third step (S300), the user's content storage amount 24 is corrected according to the user's learning history data. Referring to FIG. 6(a), the user's learning history data includes a user's forgetting curve 21, learning record 22, learning achievement rate 23, memory retention amount per content 24, and application content submission rate ( 25) and learning results (26), and extract the difficulty information (13) of the basic learning content (11) and the applied learning content (12), reflected with the data about the user's learning history By doing so, it is possible to correct the storage amount 24 of each user's content.

Looking more closely,

The user's memory retention amount per content 24 is based on the general user's average memory capacity coefficient, which can be defined as 71.52156, which can be defined as 71.52156, which is expressed as S and the memory retention amount for each general content expressed by R. ,

In the general oblivion curve defined as, reflecting the user's learning results (26),

Can be defined as The R'refers to the user's content storage amount 24, and the S'refers to the corrected memory coefficient. The corrected memory coefficient is

는 정답개수를 의미하고,

은 문제개수를 의미한다. 상기 정답개수는 사용자에게 제공된 문제개수를 사용자가 풀고난 후에 채점을 통한 정답개수를 의미하며, 상기 기본 학습 콘텐츠와 상기 응용 학습 콘텐츠를 포함하며, 상기 문제개수는 본 발명의 일 실시 예에 따라 제공한, 상기 기본 학습 콘텐츠와 상기 응용 학습 콘텐츠의 총 합을 의미한다. 여기서 상기 보정된 기억력 계수는

로 표현하는 상기 기억력 보정계수로 표현가능하며, Can be defined as

Means the number of correct answers,

Means the number of problems. The number of correct answers means the number of correct answers through scoring after the user solves the number of questions provided to the user, and includes the basic learning content and the applied learning content, and the number of questions is provided according to an embodiment of the present invention In addition, it means the total sum of the basic learning content and the applied learning content. Where the corrected memory coefficient is

It can be expressed by the memory correction coefficient represented by

는 기억력 오차를 의미하며, 상기 기억력 오차는 Can be defined as

Means memory error, and the memory error is

는 기대정답개수를 의미하며,Can be expressed as remind

Means the number of correct answers,

Can be defined as The user's application content submission rate (25) is

는 응용 콘텐츠 정답개수를 의미하고,

은 응용 콘텐츠 개수를 의미한다. 상기 응용 콘텐츠 개수는 본 발명의 일 실시 예에 따라 제공한 전체 학습 콘텐츠 중에 상기 응용 학습 콘텐츠의 수를 의미하며, 상기 응용 콘텐츠 정답개수는 제공된 상기 응용 학습 콘텐츠를 사용자가 풀고난 후 채점을 통한 정답개수를 의미한다. It is defined as, the M'refers to the user's application content submission rate 25, the M refers to the general application content submission rate,

Means the number of correct answers for application content,

Means the number of application contents. The number of application contents means the number of the application learning contents among all the learning contents provided according to an embodiment of the present invention, and the number of correct answers to the application contents is correct through scoring after the user solves the provided application learning contents It means the number.

In this case, when the learning achievement rate 23 of the user's application content submission rate 25 is low, not providing the applied learning content 12 may increase the learning effect. Therefore, the general application content submission rate varies according to the learning achievement rate 23, and when the learning achievement rate 23 is divided into three or more grades,

It may be defined as, and the user's learning result 26 may correct the user's memory storage amount 24 for each content and the user's application content submission rate 25. In addition, since the memory retention amount 24 for each user's content and the corrected memory coefficient and time constitute the forgetting curve 21, the user's forgetting curve according to the corrected memory retention amount for each user's content 24 A step of correcting (21) can be added.

In the fourth step (S400), the basic learning content 11 is extracted and arranged in the order in which the user's storage capacity for each content is low. By repeatedly providing content having a low memory storage amount 24 for each content of the user, the learning effect can be increased to the user.

The fifth step (S500) extracts the user's application content submission rate (25). The user's application content submission rate (25) is

는 응용 콘텐츠 정답개수를 의미하고,

은 응용 콘텐츠 개수를 의미하고,

은 기대정답개수를 의미하며,It may be defined as, in the formula M'refers to the user's application content submission rate, M refers to the general application content submission rate,

Means the number of correct answers for application content,

Means the number of application contents,

Means the number of correct answers,

The expected number of correct answers is

Can be defined by

In the sixth step (S600), the user's application content submission rate 25 is corrected according to the user's learning history data. As shown in Fig. 6(b), the expected content of the user's application content can be corrected by extracting the expected number of correct answers, the number of application contents, and the number of correct answers. The user's application content submission rate (25) is

는 응용 콘텐츠 정답개수를 의미하고,

은 응용 콘텐츠 개수를 의미하고,

은 기대정답개수를 의미하며,It may be defined as, in the formula M'refers to the user's application content submission rate, M refers to the general application content submission rate,

Means the number of correct answers for application content,

Means the number of application contents,

Means the number of correct answers,

The expected number of correct answers is

Can be defined by

At this time, since the application content submission rate 25 of the user does not present the application learning content 12 when the learning achievement rate 23 is low, the learning effect may be increased. When the general application content submission rate is different, and the learning achievement rate 23 is divided into three or more grades,

It can be defined as, it can be corrected by at least one of the learning result 26 or the learning achievement level of the user.

In the seventh step (S700 ), the basic learning content 11 arranged in the order of the lowest storage amount 24 for each content is converted into the applied learning content 12 according to the user's application content submission rate 25.

The eighth step S800 provides the arranged basic learning content 11 and applied learning content 12 to the user's terminal. The method of providing includes all of the components capable of communicating from the user's terminal, and in the method of communication, any method capable of communicating, such as wired communication or wireless communication, can be used.

The ninth step (S900) includes receiving the learning result 26 for the basic learning content 11 and the applied learning content 12 provided to the user's terminal in the eighth step S800, The difficulty level of the basic learning content 11 and the applied learning content 12 may be corrected according to the learning result 26.

With respect to the specific method of the method for providing user-customized learning content using the artificial intelligence, it is obvious from the standpoint of a person skilled in the art that the method implemented in the system for providing user-customized learning content using the above-described artificial intelligence may be dedicated Do.

Referring to FIG. 7, referring to an embodiment of the user's terminal 200 and the user-customized learning content providing system 100, the server 100 receives a learning content call signal from the user's terminal 200 and , The server 100 may extract and analyze data related to the user's learning history, and extract customized learning content suitable for the user. When the server 100 provides the extracted customized learning content to the terminal 200, the user solves or provides an answer to the customized learning content provided to the terminal 200. The server 100 may receive the user's answer, score it, and generate a learning result. The server 100 may transmit the generated learning result 26 to the user's terminal 200 to feedback the learning result to the user, and using the learning result 26, the user's learning history Data relating to may be corrected, and the difficulty of the learning content provided to the user may be corrected and stored as corrected data. It is possible to provide a system and method for providing customized learning content suitable for a user and improving learning efficiency through data on the user's learning history that is continuously corrected.

The disclosed content is only an example, and can be variously changed by a person having ordinary skill in the art without departing from the gist of the claims claimed in the claims, so the scope of protection of the disclosed content is specified above. It is not limited to the embodiment.

10: content management department
11: Basic Learning Content
12: Applied Learning Content
13: Difficulty by content
20: learning history big data management department
21: Oblivion curve
22: Study Date
23: Learning achievement rate
24: memory retention by content
25: application content submission rate
26: learning result
30: learning optimization engine
40: content submission department
50: content scoring department
60: content difficulty correction unit
100: user-specific learning content providing system

Claims (14)

In the learning content providing system using artificial intelligence,
A content management unit that stores data related to the learning content;
A learning history big data management unit that stores data on a user's learning history;
A learning optimization engine unit that calculates and corrects a memory retention amount for each user's content according to data related to the learning contents stored in the content management unit and data related to the learning history of the user stored in the learning history big data management unit;
A content presentation unit that provides a user with a memory storage amount in descending order based on the memory storage amount for each content calculated and corrected by the learning optimization engine unit;
A content scoring unit for scoring a user's learning result for the presented learning content;
It includes; content difficulty correction unit for correcting the difficulty of the learning content according to the learning result of the user; and
Data related to the learning content stored in the content management unit
Includes basic learning content, applied learning content and content-specific difficulty,
The data related to the learning history of the user stored in the learning history big data management unit
It includes each user's forgetting curve, learning date, learning achievement rate, memory retention amount by content, application content submission rate, and learning result.
The forgetting curve of the user stored in the learning history big data management unit,

In the general forgetting curve defined by, R means the memory retention amount for each general content, t means the time, S means the general memory coefficient, and in the general forgetting curve, the user's learning result is reflected,
R'refers to the amount of memory held by each user's content,

It is defined as, S'means the corrected memory coefficient, and the corrected memory coefficient is

Is defined as,

Means the number of correct answers,

Means the number of problems,
The user's application content submission rate is

It is defined as, M'refers to the user's application content submission rate, M refers to the general application content submission rate,

Means the number of correct answers for application content,

Means the number of application contents,

Means the number of correct answers,
The expected number of correct answers (E _c ) is

It is defined as, according to the learning progress of the user, the user-provided learning content providing system using artificial intelligence, characterized in that the provided content changes.
delete The method according to claim 1,
The content submission unit preferentially extracts the basic learning content having a low memory retention amount for each user's content, and converts the extracted basic learning content into the applied learning content according to the user's application content submission rate. A system for providing user-customized learning content using artificial intelligence, wherein the application content submission rate is corrected based on the learning result.
The method according to claim 1 or claim 3,
In the learning optimization engine unit, according to the learning result stored in the learning history big data management unit, the memory storage amount for each content is corrected, and the forgetting curve of each user is corrected according to the corrected memory storage amount for each content. User-specific learning content providing system using artificial intelligence.
delete The method according to claim 3,
The content submitting unit preferentially extracts the basic learning content having a low storage amount for each user's content, and converts the content into the applied learning content according to the user's application content submission rate expressed as M',
The user's application content submission rate is

It is defined as, M'refers to the user's application content submission rate, M refers to the general application content submission rate,

Means the number of correct answers for application content,

Means the number of application contents,

Means the number of correct answers,
The expected number of correct answers is

User defined learning content providing system using artificial intelligence, which is defined by and corrects the application content submission rate according to the learning result.
The method according to claim 6,
In the learning optimization engine unit,

It is defined as, wherein R'means the user's storage capacity for each content, t is the time, and S'is the corrected memory coefficient of the user's oblivion curve according to the learning result, A system for providing customized learning content using artificial intelligence, wherein the user's memory retention amount is corrected for each content, and the forgetting curve of the user is corrected according to the corrected memory retention amount for each user's content.
In the method for providing customized learning content using artificial intelligence,
A first step of receiving a learning content call from a user's terminal;
A second step of extracting a user's memory retention amount for each content;
A third step of correcting the user's memory retention amount for each content according to the user's learning history data;
A fourth step of extracting and arranging the basic learning contents in the order of lower storage amount of each user's contents;
A fifth step of extracting a user's application content submission rate;
A sixth step of correcting the user's application content submission rate according to the user's learning history data;
A seventh step of converting the basic learning content arranged in the order of the lowest storage amount for each content into applied learning content according to the user's application content submission rate; And
And an eighth step of providing the arranged basic learning content and applied learning content to a user's terminal.
In the third step, the data regarding the user's learning history includes the user's forgetting curve, learning date, learning achievement rate, memory retention amount for each content, application content submission rate, and learning result, and the basic learning content and the applied learning content Difficulty information is extracted and reflected together with data related to the user's learning history to correct the user's memory retention for each content.
The amount of memory held by each user's content is

In the general forgetting curve defined by, R means the memory retention amount for each general content, t means the time, S means the general memory coefficient, and in the general forgetting curve, the user's learning result is reflected,

It is defined as, R'means the user's content storage capacity, S'means the corrected memory coefficient, and the corrected memory coefficient is

Is defined as,

Means the number of correct answers,

Means the number of problems,
The user's application content submission rate is

It is defined as, M'refers to the user's application content submission rate, M refers to the general application content submission rate,

Means the number of correct answers for application content,

Means the number of application contents,

Means the number of correct answers,
The expected number of correct answers (E _c ) is

It is defined as, according to the learning progress of the user, the method of providing user-customized learning content using artificial intelligence, characterized in that the provided content changes.

The method according to claim 8,
And receiving learning results for the basic learning content and the applied learning content provided to a user's terminal in the eighth step, and correcting the difficulty of the basic learning content and the applied learning content according to the learning results. Method for providing user-specific learning content using artificial intelligence, comprising the ninth step.
delete The method according to claim 8,
The third step comprises the step of modifying the oblivion curve of the user according to the corrected user's storage capacity for each content, providing a method for providing customized learning content using artificial intelligence.
delete The method according to claim 8,
In the sixth step, the user's application content submission rate is

It is defined as, M'refers to the user's application content submission rate, M refers to the general application content submission rate,

Means the number of correct answers for application content,

Means the number of application contents,

Means the number of correct answers,
The expected number of correct answers is

It is defined by, and provides the application learning content according to the user's application content submission rate, the user tailored using artificial intelligence, characterized in that to correct the application content submission rate by the learning result received in the ninth step How to provide learning content.
The method according to claim 11,

It is defined as, wherein R'refers to the user's storage capacity for each content, t is a time, and S'is the user's forgetting curve, which means the corrected memory coefficient, for each user's content A method for providing user-customized learning content using artificial intelligence, wherein the user's forgetting curve is corrected according to the amount of possession.

Images