WO2019066115A1

WO2019066115A1 - Method for organizing knowledge system diagram of curriculum for educational platform and device for providing same

Info

Publication number: WO2019066115A1
Application number: PCT/KR2017/011067
Authority: WO
Inventors: 고범석; 서정훈; 김정민
Original assignee: (주)자이네스
Priority date: 2017-09-29
Filing date: 2017-09-29
Publication date: 2019-04-04

Abstract

Disclosed are a method for organizing a knowledge system diagram of curriculum for an educational platform, which is educational content that can be recommended for learning through a knowledge system diagram of curriculum in the field of education, and an apparatus for providing the same. The present invention enables an integrated analysis of learning data, and makes it possible to provide a recommendation of a customized learning roadmap and content by means of artificial intelligence technology. Accordingly, the present invention can provide a customized artificial intelligence standard educational platform capable of an integrated analysis of existing content.

Description

The course knowledge system for the education platform is structured and the device that provides it

The present invention relates to a method of structuring a curriculum knowledge system for an educational platform and an apparatus for providing the same, and more particularly, to an educational content that can be recommended for learning through a knowledge system diagram of a curriculum in the field of education.

Recently, traditional education methods are continuously changing according to the development of smart learning and the paradigm of education which is changing with learner - centered. Flipping learning, which is student - led in the center of professions, has been a topic of recent education.

Flipping learning, which has been attracting attention from major industrialized countries in the United States in recent years and in recent years in 2010, allows students to preview classroom lectures before class, and to improve class participation and involvement in classroom activities.

The development of this blended learning has been steadily continuing, but it has not been completed yet.

It is difficult to realize effective blended learning due to the difficulty of analyzing such prior learning activities. Therefore, it is used only as auxiliary data of traditional class.

Korean Patent Registration Nos. 100877583, 101228816, 100978091, and 101030577 have been developed as teaching methods as described above, but they have not been completed yet.

Accordingly, the present invention for solving the above-mentioned problems provides a customized artificial intelligence education standard platform capable of integrally analyzing existing contents by a standard interlocking method, a method for structuring a curriculum knowledge system for an educational platform, and an apparatus for providing the same .

Accordingly, the course knowledge system for the education platform for achieving the above object is also performed in the server for providing the educational contents according to the structuring method, and includes a problem, a step of collecting a curriculum commentary, a step of extracting a key word from the problem, Determining unit knowledge from key words in the curriculum information, determining degree of relevance between unit knowledge, and linking and storing connection information between unit knowledge.

In addition, the server for providing the problem according to the curriculum knowledge system diagram is a system in which a unit of knowledge is derived from a problem, a curriculum commentary, a database in which a knowledge diagram is stored, and a problem stored in the database, And an information extraction and conversion unit for connecting to the connection information and storing the connection information in the database.

Therefore, according to the present invention, it is possible to provide a problem matched to subsequent unit knowledge according to the unit knowledge matched to the problem and the connection information connected to the unit knowledge, with respect to the result of the problem solving by the user.

In addition, it is easy for the learner to manage the learner 's curriculum for the teacher, through the solving of the learner' s problem and feedback.

In addition, integrated analysis of learning data is enabled, and artificial intelligence technology enables personalized learning roadmap and recommendation of contents, thereby providing customized artificial intelligence education standard platform capable of integrated analysis of existing contents.

1 is a conceptual diagram of a training platform according to an embodiment of the present invention.

2 is a conceptual diagram illustrating a knowledge system according to an embodiment of the present invention.

FIG. 3 is a schematic diagram illustrating the provision of a real-time knowledge achievement evaluation model and personalized personal learning using the knowledge map according to an embodiment of the present invention.

4 is a block diagram of an AI STEM education platform according to an embodiment of the present invention.

FIG. 5 illustrates reusable modeling of content according to an embodiment of the present invention.

6 illustrates a relationship between a server and a browser according to an embodiment of the present invention.

7 shows an exemplary view of a tracking service according to an embodiment of the present invention.

FIG. 8 shows a configuration diagram of a customized artificial intelligence STEM training platform according to an embodiment of the present invention.

FIG. 9 shows evaluation model data for customized personal learning according to an embodiment of the present invention.

10 is a block diagram of a big data platform according to an embodiment of the present invention.

11 shows a big data analysis engine according to an embodiment of the present invention.

12 shows a blended learning model according to an embodiment of the present invention.

FIG. 13 shows feedback of a result of a big data analysis according to an embodiment of the present invention.

FIG. 14 is a diagram of a course knowledge system according to an embodiment of the present invention.

15 illustrates a problem providing system according to an embodiment of the present invention.

FIG. 16 conceptually illustrates the structuring of the knowledge map according to an embodiment of the present invention.

17 is a flowchart illustrating a structuring procedure of a knowledge map according to an embodiment of the present invention.

FIG. 18 conceptually illustrates creation of a knowledge map according to an embodiment of the present invention.

Hereinafter, the present invention will be described in more detail with reference to the accompanying drawings.

It is to be understood that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting, but is only limited by the scope of the appended claims. All technical and scientific terms used herein have the same meaning as commonly understood to one of ordinary skill in the art unless otherwise stated. On the contrary, the various embodiments of the present invention can be combined with any other embodiments as long as there is no clear counterpoint. In particular, any feature that is desirable or advantageous may be combined with certain other features that are desirable or advantageous.

According to an embodiment of the present invention, integrated analysis of learning data can be performed by providing a standard interworking method of fragmented educational contents, and a personalized learning roadmap and contents recommendation can be made by artificial intelligence technology.

According to an embodiment of the present invention, a method for structuring a curriculum knowledge system for an educational platform and a device for providing the same are provided.

As shown, the present invention can build a personalized AI training system based on Big Data and machine learning. In one embodiment of the present invention, a knowledge map for analyzing integrated courses is constructed. In addition, knowledge system can derive influences on academic achievement.

As shown, the knowledge map constructs a unit knowledge pool of a tree structure, and a knowledge map is created from the unit knowledge pool.

More specifically, it is possible to construct a knowledge system diagram in which the relationship between unit knowledge is dated in the form of a network model by utilizing the unit knowledge structure of the subject.

This knowledge map can be used as a basis for learning data integration analysis by deriving various variables affecting academic achievement such as learning behavior pattern, difficulty of concept, importance, and discrimination power.

Ultimately, it can be applied to a customized training platform based on machine learning through optimized knowledge system for integrated education of the subject.

As shown in the figure, it is possible to provide personalized artificial intelligence personalized learning through integration of educational contents standards. In this context, SCORM (Sharable Content Object Relation Model), which constitutes the largest proportion of domestic standardization for interworking existing contents, is a content development standard model that executes learning courses along with content composition, delivery, and search methods, How to confirm) standards can be applied.

Also, it can be applied to a scalable large data processing platform by storing and analyzing learning data through a machine learning based real - time knowledge achievement evaluation and prediction model. At this time, a big data platform based on Mpp can be constructed.

As a recommendation algorithm for artificial intelligence students' customized class, CB (Content Based), CF (Collaborative Filtering) based problem and learning recommendation algorithm can be installed.

The present invention makes it possible to interwork various devices to optimize hybrid application.

The present inventors have found that if the existing technology-based learning is to deliver popularized learning contents to a hypothetical learner group, the future direction of the learning support may be a learner The goal of the system is to optimize and implement individualized education tailored to individual characteristics. In spite of active domestic standardization efforts, application of standardization in domestic e-learning solutions and content development has not yet been applied. Standard technology also applies different standards, and it is confirmed that study on standardized interworking method to integrate them is necessary.

Therefore, it was suggested that students should be able to improve their students' academic ability by recommending customized learning to students through the evaluation of knowledge achievement and development of predictive models, and to provide teachers with real-time integrated analysis data for immediate use in blended learning.

The present invention develops learning contents through the fusion of the big data technology and the deep learning technology, and makes it possible to lay the foundation of the future basic technology which can be expanded without limitation of industrial fields such as business prediction field and control field due to the nature of the technology.

Many studies, such as existing learning outcomes measurement, are concentrated in adult and college education. Although the importance and necessity of blended learning is increasing due to the transition of education paradigm in Korea, the reason why it is not activated in elementary, middle and high schools is also here. It is urgent to develop educational standard platform that can easily introduce blended learning class. Through the present invention, students are provided with customized learning through evaluation and prediction of real-time knowledge achievement, and analytical data on the achievement of real-time students are provided to the teacher so that classes and activities suitable for the class can be prepared. Recently, the STEM (Science, Technology, Engineering, Mathematics) started to be used by the US Science Foundation is now the core of education reform in major developed countries. In order to enable this integrated STEM education through the educational platform, a knowledge system that integrates the subjects of mathematics and science was established.

Therefore, according to one embodiment of the present invention, a knowledge map capable of integrated analysis of STEM subjects can be constructed and used to derive influencing variables of academic achievement.

It may also provide a real-time knowledge achievement assessment model and a customized, self-paced education platform. According to an embodiment of the present invention, a machine learning based real-time knowledge achievement evaluation and prediction model, an artificial intelligent student customized recommendation algorithm, and an LMS / LCMS / openAPI based on an educational content standard interworking method can be installed.

According to one embodiment of the present invention, the following main functions are: a TREE structure for STEM subject unit knowledge; a knowledge system in which a relationship between unit knowledge is stored in a network model form; and a NewSQL Database (CRUD), a machine learning based real-time knowledge achievement assessment and prediction system, a personalized learning recommendation function for improving individual student academic achievement, and a big data system for evaluation, prediction and learning recommendation results. An e-learning function that provides optimized learning materials, and a standard interoperable open API that enables interworking analysis of existing contents can be implemented.

STEM (Mathematics, Chemistry, Physics) knowledge system diagram: 20,000 or more, real-time processing capability: 30,000 TPS of performance target, and Big data query speed: Less than 5USD by QPH of TPC-H as index Lt; / RTI >

In addition, it will be possible to achieve API interoperability with education companies and public institutions, with more than 90% accuracy of prediction model and 20% or more improvement of student ability (achievement).

More specifically, the present invention creates a diagram of STEM subjects (mathematics, chemistry, physics) knowledge.

In the present invention, the knowledge system diagrams the unit knowledge, which is the minimum unit of the STEM course, as a tree structure of the concept by the curriculum and the Wikipedia, collects and produces the knowledge database corresponding to it, and database it in the form of the network model. Means the corresponding educational contents.

In the US, Knowton, the highest level of education, 12,000 knowledge systems are being built. In the present invention, 20,000 STEM subject matter (mathematics, chemistry, physics) knowledge systems can be created.

In order to evaluate the knowledge achievement, the accuracy of the prediction model should be more than 90%.

Specifically, it means accuracy by verification methods of Type I and II errors by statistical hypothesis verification theory based on verification data after completion of learning of recommendation system. Learning data consists of Positive learning and Negative learning, and learning data verification is performed through Positive and Negative behavior pattern analysis (Negative). Type I Error: False Positive: It is a case that it is judged that the learning has been done properly but the learning is insufficient. Type II Error: False Negative, and the case where it is judged that the knowledge acquisition of the missing student is completed.

The DB processing speed of the Big Data Platform means the DB performance of the TPC-H Benchmark method. TPC-H modeled the situation in which impromptu query and modification transactions of several users in a decision model such as data warehouse are performed in table and data is input batch from OLTP system to database of pseudo-support decision system. Specifically, the system is evaluated by the query processing index QphH per hour, and the investment cost per query divided by the construction cost of the system performing the QphH test. At this time, the equation for calculating the evaluation index of QphH is shown in Equation 1 below.

Here, QI (i) denotes the execution time of the TPC-H query, n is the number of evaluation subject queries, SF is the evaluation data size, and the unit is GB. In the case of overseas, the TPC-H evaluation value shows a high performance of less than 3 USD. However, this is a very high system construction cost, which is difficult to implement in medium and small companies other than large companies. In the case of medium and small enterprises, most of the domestic and foreign companies are calculating their performance in the form of appliances. On average, the figures are below 5 ~ 10 USD.

For the high availability of the Big Data Platform In a distributed system, when a node reaches the Fail state, that is, it becomes inoperable, it is restored without affecting the performance of the entire system, in

It means that it is restored normally. That is, it may mean that each node quickly restores the failure in the failure state to provide a smooth service.

In the distributed database system, failure recovery performance is judged according to the failure scenario for each node, and most of the features of the distributed system are characterized by smooth driving. Therefore, the recovery level is provided within 5 to 10 seconds. Therefore, it is desirable to restore within a maximum of 5 seconds according to the scenario for each node.

(Group 2) of the learning group (group 1) learned based on the platform of the present invention with respect to the degree of achievement improvement of the student's academic ability.

Specifically, the highest level in the field, the United States, knewton, 12.5% of the academic achievement of the improvement is shown.

Based on the results, the learning curriculum of NCIC (National Curriculum Information Center) and the achievement standard of the national curriculum were prepared during the learning experiment, and the learning group (group 1) and the control group (group 2) Through the evaluation of the end of the follow-up verification, the achievement rate of the achievement through this task should reach more than 20%.

APIs can be linked with educational companies and public institutions. In the present invention, this means that the learning contents (problem, knowledge) of the existing e-learning learning site are matched through API interoperability with the knowledge database of the platform. In order to evaluate with the knowledge database constructed in the platform of the present invention and to link the learning contents, the existing education company and the e-learning site operated by each trial can be linked with API more than two places.

As shown in the figure, the framework for the STEM educational platform for student-based artificial intelligence based on machine learning and the framework for integrated learning data analysis based on machine learning based on the existing contents are linked with the API system.

Specifically, it is possible to integrate the existing developed contents with the existing LCMS through legacy interface design, and to analyze the contents and LCMS / LMS relation for SCORM application. At this time, a content model capable of reusing contents can be used.

As shown, the SCORM standard extends the platform-optimized Meda-data design and implements communication between the browser and the server to implement the LMS that conforms to the SCORM standard.

As shown, the SCORM-based Learning Contents Management System (LCMS) customizes the escubeDCMS (Digital Contents Management System) by applying the SCORM standard. Contents Management and Management, Category Management, Deploy Management, Contents Statistics Management, and SCORM-based Learning Management System (LMS) can be installed. In addition, evaluation data such as learning activity (meaningful tracking data) and incorrect response rate data such as interworking of AI based learning recommendation algorithm, real - time knowledge achievement evaluation and prediction model applying technology, tracking data service for big data formation for learning data analysis, The degree of difficulty, importance, and discrimination of the contents can be confirmed.

As shown in the figure, learning support services such as learning, task, and evaluation are implemented, teaching support is implemented, classes are opened, student management and evaluation functions are implemented, and management support, user management ), Curriculum management, learning management, and course management.

As shown in the figure, the present invention is a system for providing customized education as an infrastructure system for building a customized artificial intelligent STEM education platform, in which a student's evaluation data, a student profile, a knowledge system based on unit knowledge, We analyze and store large data Mart that processed interrelated data using DB in big data platform which can process at high speed.

In the present invention, the data types according to types in the present invention are classified into unit knowledge information constituting knowledge map using unit knowledge, question bank information based on unit knowledge, and evaluation result data based on unit knowledge information and question bank information And one big data analysis information.

Specifically, the unit knowledge information is knowledge used in a unit of a specific subject, and it is implemented as unit knowledge information based on a wiki, unit knowledge mutual connection information, unit knowledge and problem bank connection information, and learning course connection information for unit knowledge .

The question bank information is basic item information, and can be composed of subjects, unit, difficulty level, and discrimination power. Further, the unit knowledge link information may be composed of unit knowledge related information for solving the problem, weight information per unit knowledge, and the like. In addition, the analysis information can be composed of positive error analysis information according to clustering, and analysis information of positive / negative according to learning type / course.

Big data analysis information is analysis information according to unit knowledge and question type of question bank. Specifically, the association analysis information includes association analysis information between a problem and a problem, and association analysis information between a problem and a unit knowledge. In addition, the clustering analysis information includes problem clustering analysis information according to the learning level, and learning class clustering analysis information according to the difficulty level and the discrimination power. Also, the connection analysis information can be made of learning course information using a directional graph.

In addition, evaluation model data for customized personal learning is constructed.

In order to construct an evaluation model for customized personal learning according to the present invention, a question based on unit knowledge is constructed, an assessment composed of problems of the question bank is constructed, and a study course course.

As shown, the problem is one or more unit knowledge information items such as a subject, a unit, a difficulty level, a discrimination power, and item information.

The evaluation consists of evaluation items on unit knowledge according to the learner 's grade. In addition, CB (Contents-Based) -based recommendation algorithm can be applied to evaluation item recommendation before pre-learning. In addition, recommendation items based on Deep Belief Network (DBN) and Collaborate Filtering (CFN) can be applied to the evaluation items after the preliminary learning.

The learning course constitutes a unit knowledge learning process according to the learner 's class. In addition, CB recommendation algorithm is applied to recommendation course before pre-learning. In addition, recommendation of learning course after the pre-learning can apply RNN (Recurrent Neural Network) -CF based recommendation algorithm.

Meanwhile, according to an embodiment of the present invention, a big data platform capable of processing large-capacity data reconstructed in association with evaluation result data for customized learning recommendation at high speed is based. The Big Data Platform consists of a data collection / storage / analysis subsystem and consists of a GUI-based workflow for the data processing flow.

As a data collection subsystem, it uses Apache Kafka, a message queue method, to collect reliable data. As a data storage subsystem, it supports semi-structured data storage of HDFS and JSON format for storing unstructured data such as unstructured / I use NoSQL (MongoDB). We use MPP-based distributed D / W (KHIRON) for large-scale formal data processing as the formal data.

The data analysis sub-system implements an artificial intelligence-based learning recommendation system, a content-based recommendation system, and a collaborative filtering recommendation system based on an unstructured data-based data analysis system (Mahout). In addition, a formal data-based data analysis system and Clustering, Classification, Association, and Rank can be used.

The data workflow subsystem can also be provided by a GUI-based process-specific workflow and a Pentaho-based Plug-In method.

As shown, according to an embodiment of the present invention, for a system capable of processing large-capacity data reconstructed in association with evaluation result data at high speed for customized learning recommendation, MPP-based large data using Khiron, which is a big data solution, Implement the platform. Khiron is a new database that can use both RDB and NoSQL. It is designed in MPP-based 3-tier configuration, and uses a solution that is easy to expand and load-balance and ensures high availability. In the present invention, it can be used as a core solution for linking with a processed data loading and analysis engine, and linking with a recommendation engine.

According to one embodiment of the present invention, in order to provide a customized education, data of students' evaluation result, a student profile, a knowledge system based on unit knowledge, and a large-capacity data Mart is analyzed at high speed.

The present invention can be divided into clustering analysis and association analysis. The clustering analysis classifies the students according to the problem difficulty, according to the problem. The association analysis analyzes information based on the recommendation based on the relation between the problem and the problem, and the information related to the recommendation based on the association between the problem and the unit knowledge.

As a recommendation engine for a customized learning system, a machine learning based recommendation system is implemented to evaluate the student's current academic achievement level in real time based on past learning data. A content-based recommendation system for generating the prior learning data, a collaborative filtering recommendation system based on the prior learning data, a problem recommendation system for evaluating the academic achievement level, and a personalized personal learning Implement a course recommendation system that designs progress.

In accordance with an embodiment of the present invention, unit knowledge subdivision and tree-type structuring of one subject of STEM subject in middle and high school are structured and a structured unit knowledge base knowledge DB is constructed. Specifically, based on the curriculum of NCIC (National Curriculum Information Center) of the Ministry of Education and the open knowledge of Wikipedia, unit knowledge can be structured and a knowledge database can be collected and produced by building knowledge pool.

According to one embodiment of the present invention, a Learning Contents Management System (LCMS) suitable for hybrid learning is a technology capable of reusing contents generated by a teacher and a learner, and Meta-data for enhancing content reusability. Contents Packaging system and meta-data management system can be applied. It also provides a tracking service for on / off integrated learning data analysis.

We implement a blended learning environment through LMS (Learning Management System) suitable for mixed learning, AI customized learning recommendation algorithm, Testing / Assessment System, and on / off integrated learning evaluation system. It provides a teaching service suitable for mixed learning. It is desirable to have flexibility by developing Course redesign that instructor can design course directly, and it can provide real-time / non-real-time feedback through cooperative learning monitoring. To enable collaborative learning through the development of the Team Building function, and to enable discussions and inquiries such as teacher-student mutual function to be evaluated. (Digital Contents Billing System) for platform commercialization service and customized escubeDCRS (Digital Contents Re-packing-sales System).

As shown in the figure, according to the embodiment of the present invention, the analysis result data calculated using the Big Data Analysis Platform is composed of problem association analysis information, problem and learner clustering analysis information, and problem connection analysis information. By analyzing the unit information constituting each information by using the above three pieces of information, a variable that affects the academic achievement is derived. The present invention may be further applied to a learning recommendation system using derived learning achievement variables. Specifically, the problem recommendation system can be achieved through problem association analysis and problem and learner clustering analysis. In addition, the course recommendation system can be done through problem connection analysis and problem and learner clustering analysis.

In the present invention, the course knowledge system chart is for a training platform, and can be installed in a computer such as a server or a client for providing it to a user.

The knowledge system can define simple knowledge composed of a number of knowledge cells as a knowledge mesh network connected with a knowledge synapse like a human brain structure. Furthermore, analysis of learning data is possible based on this.

The Knowledge Cell is a minimum unit that enables students to measure the knowledge level of each student through the problem solving based on the knowledge cell rather than the center of the existing content lecture.

It can provide students with the learning they need, rather than providing the same lectures and problems to students at various levels.

Such a knowledge cell expresses the level of understanding and acquisition of knowledge and the order of acquisition according to the connection strength of Knowledge Snaps, and finds problems of specific knowledge according to the logical operation of knowledge. In this way, specific knowledge with weak strength can be found from the specific connection relation of the problematic part.

As shown in the figure, a system as shown in FIG. 15 is required to implement the knowledge map of the present invention. That is, the server 10 may be divided into a server 10 for providing information and processing information according to the knowledge map, and a client 20 for learning or correcting information stored in the server according to the actual knowledge map.

The information extracted from the problem database (BD) of the server 10, the unit knowledge of the problem, the unit knowledge, and the curriculum information is converted into a web page by the information extraction / conversion unit 14, Such as the Internet. The learning information display unit displays the problem and the learning list on an output device such as a PC screen or a screen of a mobile device. Specifically, if a learning process is selected from the learning list, a problem mapped to the learning process is provided so that the user can solve the problem.

Here, when the user solves the problem, the related problem can be displayed according to the connection information matched to the unit knowledge of the problem (for example, essential learning, selective learning, reference learning).

Also, a third user who is aware of the contents of learner, lecturer, or learning object can use the client's information editing section to view the problem and learning list displayed on the learning information display section, Do.

In one embodiment of the present invention, the server may include a problem DB 11, a training course description DB 12, a knowledge base DB 13, and an information extraction and conversion unit 14, An information editing unit 22, a learning list processing unit 23, and a learning information display unit 24, as shown in FIG.

Hereinafter, the structuring of the knowledge system diagram of the present invention will be described in more detail.

According to the present invention, the structuring of the knowledge map is performed in a server for providing educational contents, and includes collecting a problem and a curriculum description book (S100), extracting a key word from a question (S200) (S300) of determining unit knowledge from key words in the information, a step (S400) of determining an association degree between unit knowledge, and a step of connecting and storing connection information between unit knowledge (S500) .

FIG. 16 conceptually illustrates the structure of a knowledge map according to an embodiment of the present invention, and FIG. 17 is a flowchart illustrating a structure of a knowledge map according to an embodiment of the present invention.

16 and 17, the knowledge structure of the present invention is structured based on information input by a user to a client or a server, and further provides a problem to a client according to the structured knowledge diagram .

More specifically, a training course guidebook and a sharing problem (hereinafter referred to as " problem ") input by the user can be collected and stored in the database DB of the present invention. After deriving the connection information between the unit knowledge, structure the knowledge map by connecting the unit knowledge to the connection information. Each unit knowledge is then matched to each problem. In this way, the information extraction and conversion part of the server extracts the unit knowledge and the connection information based on the problem of the database and the instruction manual of the training course, and then the knowledge system is also processed and stored in the DB.

As shown in FIG. 17, first, the server collects the problem and the training course description and stores the problem DB and the training course description DB in the knowledge structure of the present invention. From the problem database, we extract key words, for example nouns, from the words that constitute the problem. In other words, the key word is to extract words separated by words from the problems stored in the database.

And we find the same curriculum as the key word in the problem from the curriculum information made up of the text of the curriculum commentary. At this time, whether or not the same curriculum is to be found can be determined from whether or not synonyms previously linked from the same or the same word in the word are included in sentences or phrases. A program for implementing such a function may be installed, but a known technique may be used. At this time, it is desirable to search for and search for the precedence and following of the same curriculum together with the similar curriculum.

Thus, unit knowledge is determined from key words in curriculum information. see. For example, in the case of mathematical education, the unit knowledge is determined by a series of related knowledge concepts such as a square root, a square root of a positive number, a square root of a negative square, a property of a square root, In addition, in the field of chemistry education, the concept of chemical problems such as elements, periodic table, metal elements, metals, iron (Fe), iron (Fe) Associated unit knowledge can be determined.

In other words. Among the nouns extracted from the problem, it is difficult to say that the words that are not related to education or other concepts are core words. In order to have discernment among such words, core words are extracted by combining them with the curriculum commentary, . &Lt; / RTI >

Next, the unit knowledge determined in the above can be determined as the relevance degree between them. This degree of relevance can be an important indicator to classify as essential learning, selective learning, or reference learning in which the relationship between unit knowledge is defined. In the present invention, the degree of relevance can be determined according to the order of the curriculum existing in the characters of the curriculum instruction book of the key word. More specifically, the degree of relevance between unit knowledge may be determined according to the words of the phrase in accordance with the top-down relation of the concepts or the order relation in the course of education, and sometimes a predetermined user or operator may be required to intervene. However, in the present invention, it is possible to determine the advanced relevance degree grading through the various big data processing techniques.

The degree of relevance between unit knowledge is determined by order of unit knowledge according to degree of association between unit knowledge determined by order of connection information, and in the present invention, it can be divided into three grades. Specifically, if the relationship between the unit knowledge and the simple curriculum can be divided into 1, 2, or 3 grades, the grade is determined to be 1 grade. If the relationship between the unit knowledge and the curriculum is a word The second level is determined. In addition, if the unit knowledge is in a horizontal relation that is not a top-down relationship, or in a simultaneous learning relationship that is simultaneously educated in a curriculum, it is determined as a third grade.

As shown, when the unit knowledge is determined from the problem, the connection information between the unit knowledge is set. It is preferable that the order of the unit knowledge is determined according to the degree of relevance between the unit knowledge determined according to the order of the curriculum.

Such connection information can be set from the relevance class. More specifically, in the case of the first degree to the first degree to the degree of relevance classified into the first to third classes, it is essential that any one unit knowledge should learn another unit knowledge as essential. For example, to be. This essential learning means that a second problem can be solved in order to solve a first problem.

When the degree of relevance is class 2, it is possible to solve a second problem by solving a first problem as a selective learning, and selectively solving a second problem selected in a first problem. At this time, if the first problem can not be solved, the user can select whether to solve the second problem.

Finally, if the relevance grade is grade 3, it is a second problem similar to the first problem as the reference learning. The first problem and the second problem are related to the unit knowledge, which is lower than the first and second grades, to be.

In this way, if the relevance class is a first, second, or third class, the connection information of each unit knowledge is different in the case of the relation of essential learning, selective learning, and reference learning. It can be structured. For example, the required learning of the first grade can be connected to the solid line, the second grade optional learning can be connected to the zigzag line, or the reference learning of the third grade can be connected to the dashed dotted line.

The present invention further includes a step S600 of providing a problem mapped to the unit knowledge according to a user's unit knowledge selection and continuously providing other unit knowledge problems according to the connection information depending on whether the problem is solved It is possible.

Such unit knowledge may be electronically connected, and when the unit knowledge is electronically connected differently, it is impossible to solve or solve a problem matched to the unit knowledge, It is possible to provide a problem matched to the subsequent unit knowledge according to the unit knowledge and the connection information connected to the unit knowledge.

In addition, integrated analysis of learning data is enabled, and a personalized learning roadmap and recommendation of contents can be made by artificial intelligence technology, thereby providing a customized artificial intelligence education standard platform capable of integrated analysis of existing contents.

Although the present invention has been described in detail with reference to the preferred embodiments thereof, those skilled in the art will appreciate that various modifications, additions and substitutions are possible, without departing from the scope and spirit of the invention as defined by the appended claims. It will be apparent to those skilled in the art that various changes and modifications may be made without departing from the scope of the present invention. It is therefore to be understood that the above-described embodiments are illustrative in all aspects and not restrictive. The scope of the present invention is defined by the appended claims rather than the foregoing detailed description, and all changes or modifications derived from the meaning and scope of the claims and the equivalents thereof are included in the scope of the present invention Should be interpreted.

Claims

And is performed in a server for providing educational contents,

A problem, and a course instruction book;

Extracting key words from the problem;

Determining unit knowledge from key words in curriculum information;

Determining an association degree among the unit knowledge; And

Connecting and storing connection information between unit knowledge;

A structured method of knowledge structure for curriculum for educational platform.
The method according to claim 1,

Wherein the key word is a word extracted from the problem stored in the database.
The method according to claim 1,

The relevance class is determined according to the order of the curriculum existing in the characters of the curriculum description of the key word.
The method according to claim 1,

Wherein the order of the unit knowledge is determined according to the degree of relevance between the unit knowledge determined according to the order of the curriculum.
The method according to claim 1,

Further comprising the step of providing a problem mapped to the unit knowledge according to the user's unit knowledge selection and continuously providing the problem of other unit knowledge according to the connection information according to whether or not the problem is solved, .
As a server for providing problems by the knowledge map,

Problems, a curriculum commentary, a database in which a knowledge map is stored,

And an information extraction and conversion unit for deriving unit knowledge from a problem and a curriculum description book stored in the database and connecting unit knowledge to connection information and storing the unit information in a database.