KR102004180B1 - Apparatus and method for extracting similar test problem using recognition of test paper - Google Patents

Apparatus and method for extracting similar test problem using recognition of test paper Download PDF

Info

Publication number
KR102004180B1
KR102004180B1 KR1020170152320A KR20170152320A KR102004180B1 KR 102004180 B1 KR102004180 B1 KR 102004180B1 KR 1020170152320 A KR1020170152320 A KR 1020170152320A KR 20170152320 A KR20170152320 A KR 20170152320A KR 102004180 B1 KR102004180 B1 KR 102004180B1
Authority
KR
South Korea
Prior art keywords
test paper
question
test
unit
text
Prior art date
Application number
KR1020170152320A
Other languages
Korean (ko)
Other versions
KR20180073440A (en
Inventor
김학현
Original Assignee
김학현
주식회사 유씨아이
주식회사 유씨아이크레아토
주식회사 논리수학
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 김학현, 주식회사 유씨아이, 주식회사 유씨아이크레아토, 주식회사 논리수학 filed Critical 김학현
Publication of KR20180073440A publication Critical patent/KR20180073440A/en
Application granted granted Critical
Publication of KR102004180B1 publication Critical patent/KR102004180B1/en

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/10—Services
    • G06Q50/20—Education
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
    • G06F16/33—Querying
    • G—PHYSICS
    • G09—EDUCATION; CRYPTOGRAPHY; DISPLAY; ADVERTISING; SEALS
    • G09B—EDUCATIONAL OR DEMONSTRATION APPLIANCES; APPLIANCES FOR TEACHING, OR COMMUNICATING WITH, THE BLIND, DEAF OR MUTE; MODELS; PLANETARIA; GLOBES; MAPS; DIAGRAMS
    • G09B7/00—Electrically-operated teaching apparatus or devices working with questions and answers
    • G09B7/06—Electrically-operated teaching apparatus or devices working with questions and answers of the multiple-choice answer-type, i.e. where a given question is provided with a series of answers and a choice has to be made from the answers

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Tourism & Hospitality (AREA)
  • Educational Technology (AREA)
  • Educational Administration (AREA)
  • Human Resources & Organizations (AREA)
  • Health & Medical Sciences (AREA)
  • Economics (AREA)
  • General Health & Medical Sciences (AREA)
  • Marketing (AREA)
  • Primary Health Care (AREA)
  • Strategic Management (AREA)
  • General Business, Economics & Management (AREA)
  • General Engineering & Computer Science (AREA)
  • Databases & Information Systems (AREA)
  • Data Mining & Analysis (AREA)
  • Computational Linguistics (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

The present invention compares the text information for each problem generated by extracting only the text of each problem from the pre-stored problems and the text information for each problem of the test paper generated after recognizing the test paper, and measures and stores the similarity. The present invention relates to an apparatus and a method for extracting a similar problem through test paper recognition that can be searched and extracted.
The similar problem extraction apparatus, the test question DB unit 110; A problem-specific text extraction unit 120 for extracting problem-specific text information and storing it in the test question DB unit 110; An administrator scan unit 130 for storing the scanned test paper image in the test question DB unit 110; A test paper problem separating unit 140 for processing the test paper image to separate test paper problems which are image information for each problem, and then storing the test paper images in the test question DB unit 110; A problem OCR unit 150 for extracting text for each test paper question from the test paper questions and storing the test paper text in the test question DB unit 110; A problem similarity measurer 160 which compares the test paper question text and the text information for each question and extracts a problem similarity and then extracts each compared problem as a similar problem and stores it in the test question DB unit 110; And a measurement result inquiry and selection unit 170 for retrieving similar problems having similarity for each test paper question, and selecting and extracting similar problems for problem questions.

Description

Apparatus and method for extracting similar problems through test paper recognition {APPARATUS AND METHOD FOR EXTRACTING SIMILAR TEST PROBLEM USING RECOGNITION OF TEST PAPER}

The present invention relates to extracting a similar problem, and more specifically, to the problem-specific text information generated by extracting only the problem-specific text from the pre-stored problems, and the problem-specific text information of the test paper generated after recognizing the test paper The present invention relates to a similar problem extracting apparatus and method through recognition of a test sheet, by comparing and measuring similarity, and storing and searching similar problems for each problem.

With the development of internet technology, a system for taking various tests online is being provided.

Accordingly, Korean Patent Laid-Open Publication No. 2004-0015786 discloses a method of automatically submitting a test question on the Internet that generates a problem folder by classifying problems by similar problems, and then randomly extracts a problem in the problem folder. .

However, the above-described conventional technology has a function of allowing the user to randomly extract a test question, but does not automatically extract a problem similar to the existing test question.

Republic of Korea Patent Publication No. 2004-0015786

Accordingly, the present invention is to solve the above-described problems of the prior art, when scanning and entering the existing test paper, by extracting similar problems for each problem, to create a test or self-study for the problems similar to the previous problem It is an object of the present invention to provide an apparatus and method for extracting similar problems through test strip recognition.

Similar problem extraction apparatus of the present invention for achieving the above object,

The text information for each problem, the test paper image as a scanning image of the test paper, the test paper problem which is a problem image separated from the test paper image, the test paper text extracted from the test paper problem, the text for each test paper and the text information for each question A test question DB unit 110 for storing a database of measurement results for each test question, which is information of similar problems having a similarity level;

A problem-specific text extraction unit 120 for extracting problem-specific text information and storing it in the test question DB unit 110;

An administrator scan unit 130 for storing the scanned test paper image in the test question DB unit 110;

A test paper problem separating unit 140 for processing the test paper image to separate test paper problems which are image information for each problem, and then storing the test paper images in the test question DB unit 110;

A problem OCR unit 150 for extracting text for each test paper question from the test paper questions and storing the test paper text in the test question DB unit 110;

A problem similarity measurer 160 which compares the test paper question text and the text information for each question and extracts a problem similarity and then extracts each compared problem as a similar problem and stores it in the test question DB unit 110; And

And a measurement result inquiry and selection unit 170 for retrieving similar problems having similarity for each test paper question, and selecting and extracting similar problems to solve a question.

The problem OCR unit 150,

Designate mathematical symbols and mathematical formulas as main components and numbers as variables to extract text for each test questionnaire.

The problem similarity measuring unit 160,

At least one comparison target among specific mathematical symbols, mathematical formulas, mathematical terms, and principles and contents of the problem is designated by the user, and the problem similarity is extracted based on the designated comparison target as a priority.

The object of comparison of the principle and the content of the problem is to extract the problem similarity using the machine learning engine.

The test question DB unit 110,

A problem-specific text DB 111 for storing problem-specific text information extracted by the problem-specific text extracting unit 120;

A test paper DB 113 for storing the scanned test paper image output from the manager scanning unit 130;

A test paper problem DB 115 for storing a test paper problem, which is image information for each test generated by the test paper problem separating unit 140, by image-processing the test paper image for each test paper number;

A test paper question-specific text DB 117 for storing a test paper question-specific text extracted from the test paper questions stored in the test paper question DB 115 by the problem OCR unit 150; And

A problem paper measurement result DB (119) for storing the similarity generated by the problem similarity measurement unit (160) by comparing the test paper problem text and the problem text, as the test paper problem measurement result information; It may be configured to include.

The problem-specific text information may be extracted from the problems of the problem bank and stored, and may include problems owned by a math problem DB unit.

The test paper image stored in the test paper DB 113 may be stored with school / year / grade / semester / intermediate or end-of-year information.

The test paper problem separator 140,

The test paper image may be configured to separate images based on vertical lines, margins, and question numbers, and extract the question-specific images as test paper questions.

Similar problem extraction method of the present invention for achieving the above object, the problem-specific text DB 111, the test paper DB 113 and the test paper problem DB 115, the test paper problem-specific text DB 117 and the test paper problem measurement Exam DB including the result DB (119) (110), problem-specific text extraction unit 120, administrator scan unit 130, test paper problem separation unit 140, problem OCR unit 150, problem similarity measurement In the similar problem extraction method by the similar problem extraction apparatus 100 including the unit 160 and the measurement result inquiry and selection unit 170,

A problem-specific text DB generation process of storing the problem-specific text information extracted from the pre-stored and registered problems by the problem-specific text extraction unit 120 in the problem-specific text DB 111 (S100);

A test paper DB generation process of classifying the scanned test paper image by the manager scanning unit 130 and storing it in the test paper DB 113 (S200);

A test paper problem DB generation process of generating a test paper problem that is image information for each question by the test paper problem separating unit 140 by processing the image of the test paper and storing the test paper problem DB by the test number for each test paper (S300);

A problem paper text generation process (S400) for the problem OCR unit 150 to extract the text for each test paper question from the test paper questions stored in the test paper question DB 115 and to store the text in the test paper question text DB 117;

The problem similarity measurement unit 160 compares the test paper text for each question and the text for each problem to measure similarity, and the problems having the text for each problem and the measured similarity as the test paper problem measurement result information for each test paper problem. DB 119 measurement process DB generation process stored in the DB (119) (S500); And

The measurement result inquiry and selection unit 180 searches for and outputs the similarity and similarity correspondence problem measured for each test paper problem according to the input command, and the measurement result inquiry for outputting a problem corresponding to the input selection signal as a question. Including the selection process (S600),

The text DB generation process for each test paper question (S400),

Through the problem OCR unit 150, the mathematical symbols and mathematical formulas are designated as main components and numbers are designated as variables to extract text for each test paper question,

The measurement result DB generation process for each test paper problem (S500),

A specific mathematical symbol, a mathematical construct, a word, a phrase expression pattern, and a comparison target of at least one of the principle and the content of the problem are designated by the user, and the problem similarity is extracted based on the priority of the designated comparison target,

The object of comparison of the principle and the content of the problem is to extract the problem similarity using the machine learning engine.

The test paper image stored in the test paper DB 113 may be stored with school / year / grade / semester / intermediate or end-of-year information.

The test paper problem extracted by the test paper problem separating unit 140 in the test paper problem DB generation process (S300),

The test paper image is characterized by a problem-specific image extracted by separating images based on vertical lines, margins, and problem numbers.

Similar problem extraction apparatus and method of the present invention having the above-described configuration, the problem-specific text information generated by extracting only the problem-specific text from the pre-stored problems, and the problem-specific text information of the test paper generated after recognizing the test paper In comparison, by measuring and storing the similarity, it is possible to search for and extract similar problems for each problem, thereby enabling learning about similar problems to those included in a specific test paper, thereby significantly improving the learning effect. It provides an effect that can be improved.

1 is a block diagram of a similar problem extraction apparatus 100 according to an embodiment of the present invention.
Figure 2 is a flow chart showing the processing of the similar problem extraction method of the present invention.
3 is a view showing a sequential process of the similar problem extraction method for each configuration of the present invention.
4 is a view showing the test paper problem separation criteria by the test paper problem separation unit 140.
5 is a functional block diagram of a similar problem extraction server 1 equipped with a similar problem extraction device 1;
6 is a configuration diagram of a similar problem extraction service providing system for providing a similar problem extraction service through the Internet using the similar problem extraction server 1;
7 illustrates a principle database example associated with a problem in accordance with an embodiment of the present invention.
8 illustrates a term database example associated with a problem in accordance with an embodiment of the present invention.

In the following description of the present invention, detailed descriptions of well-known functions or configurations will be omitted when it is deemed that they may unnecessarily obscure the subject matter of the present invention.

Since the embodiments according to the concept of the present invention can be variously modified and have various forms, specific embodiments will be illustrated in the drawings and described in detail in the present specification or application. However, this is not intended to limit the embodiments in accordance with the concept of the present invention to a particular disclosed form, it is to be understood that the present invention includes all changes, equivalents, and substitutes included in the spirit and scope of the present invention.

When a component is referred to as being "connected" or "connected" to another component, it may be directly connected to or connected to that other component, but it may be understood that other components may be present in between. Should be. On the other hand, when a component is said to be "directly connected" or "directly connected" to another component, it should be understood that there is no other component in between. Other expressions describing the relationship between components, such as "between" and "immediately between," or "neighboring to," and "directly neighboring to" should be interpreted as well.

The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. Singular expressions include plural expressions unless the context clearly indicates otherwise. As used herein, the terms "comprise" or "having" are intended to indicate that there is a feature, number, step, action, component, part, or combination thereof that is described, and that one or more other features or numbers are present. It should be understood that it does not exclude in advance the possibility of the presence or addition of steps, actions, components, parts or combinations thereof.

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

1 is a block diagram of a similar problem extraction apparatus 100 according to an embodiment of the present invention.

As shown in FIG. 1, the similar problem extracting apparatus 100 includes a test question DB unit 110, a text extracting unit 120 for each problem, an administrator scanning unit 130, a test paper problem separating unit 140, and a problem OCR unit 150. ), The problem similarity measurement unit 160 and the measurement result inquiry and selection unit 170 is configured.

The test question DB unit 110, the questions are text information for each problem, a test paper image as a scanning image of the test paper, a test paper problem that is a problem image separated from the test paper image, the test paper problem-specific text extracted from the test paper problem, the text of the test paper problem Database and store the measurement result for each test question, which is similar information with similarity, which is generated by comparing the text information of each problem and the problem.

To this end, the test question DB unit 110 is a problem-specific text DB (111) for storing the text information for each problem extracted by the problem-specific text extraction unit 120, the scanned test paper image output from the administrator scan unit 130 Test paper DB (113) for storing the test paper problem DB for storing the test paper problem, which is the image information for each problem generated by processing the test paper image by the test paper image separation unit 140 for each test paper, the problem OCR The test paper question-specific text DB 117 and the question likelihood measurer 160 which store the text of the test paper question extracted from the test paper questions stored in the test paper question DB 115 by the unit 150 and the text of the test paper question. And a test paper problem measurement result DB 119 for storing the similarity generated by comparing the text for each problem and the text for each problem as the test result information for each test paper problem.

The problem-specific text extraction unit 120 is configured to extract problem-specific text information and store it in the test question DB unit 110. Here, the problems may be problems held by problems of a problem bank or a math problem DB unit. At this time, the extracted text information for each problem is used for comparison for measuring similarity with the problem included in the test paper.

The manager scanning unit 130 is configured to store the scanned test paper image input through the scanning unit 60 in the test question DB unit 110. In this case, the test paper image stored in the test paper DB 113 may be stored with school / year / grade year / semester / middle or term information.

The test paper problem separator 140 is configured to image the test paper image to separate the test paper problem, which is image information for each problem, and to store the test paper DB 110. The image processing for the test paper unknown may be configured to extract the image for each test paper into test paper problems by separating the test paper image based on vertical lines, margins, and problem numbers (see FIG. 4).

The problem OCR unit 150 is configured to extract the text of the test paper problem from the test paper problems which are images of the test paper problem separated by the test paper problem separating unit 140 and store the text in the test question DB unit 110. In this case, the extracted test questions may be customized to mathematics to increase the recognition rate. That is, in the case of a mathematical problem, in the case of an expression, the mathematical symbol leading to the expression may be extracted as a main component and stored together. For example, Root (2) and Root (3) are stored as Root is extracted as the main component, 2 and 3 are treated as X variables, and both expressions use the same mathematical symbols. In addition, if the proportional expression 3: 4 = 6: X, the mathematical formula of: =: is extracted and stored as the main component. In addition, the extracted text for each test sheet problem is preferably generated so that the user can correct it.

The problem similarity measurement unit 160 compares the text of each test paper with the text information of each question, extracts a problem similarity, and then extracts each compared problem as a similar problem and stores the similar problem in the test question DB unit 110. It is composed. In this case, in order to increase the accuracy of the measured sentence similarity, machine learning may be applied to the sentence similarity measuring method.

The measurement result inquiry and selection unit 170 is configured to be able to inquire similar problems having a similarity for each test paper problem, and to select similar problems to be extracted for the question.

That is, the measurement result inquiry and selection unit 170 extracts and displays problems having similarities for each test paper problem, and provides a user interface to select a question to be asked among the displayed problems. Find similar questions by question or select questions for the exam.

In addition, the similar problem extracting apparatus 1 may further include an Internet service unit 170 that provides an interface for providing a similar problem question service through the Internet.

2 is a flowchart showing a process of the similar problem extraction method of the present invention, Figure 3 is a diagram showing a sequential process of the similar problem extraction method for each configuration of the present invention.

2 and 3, the similar problem extraction method, the problem-specific text DB 111, test paper DB 113, test paper problem DB 115, test paper problem-specific text DB 117 and test paper problem-specific measurement results Test questions DB unit 110, DB-119 including the text extraction unit 120, administrator scanning unit 130, test paper problem separation unit 140, problem OCR unit 150, problem similarity measurement unit In the similar problem extraction method by the similar problem extraction apparatus 100, including the 160 and the measurement result inquiry and selection unit 170, the problem-specific index DB generation process (S100), test paper DB generation process (S200) ), The test paper DB generation process (S300), the test paper question-specific text DB generation process (S400), the test results for each question paper DB generation process (S500) and the measurement results inquiry and selection process (S600).

In detail, the problem-specific text information is stored in the problem-specific text DB 111 by the problem-specific text extracting unit 120 stored in the problem-specific text DB generation process (S100). Create a DB 111.

In the test paper DB generation process (S200), the manager scan unit 130 classifies the scanned test paper image and stores the scanned test paper image in the test paper DB 113 to generate a test paper DB 113.

Thereafter, in the test paper problem DB generation process (S300), the test paper problem separation unit 140 processes the test paper image to generate a test paper problem, which is image information for each problem, and stores the test paper problem DB 115 for each question number. By doing so, the test paper question DB 115 is generated.

4 is a diagram illustrating a test paper problem separation criteria by the test paper problem separation unit 140.

As shown in FIG. 4, the test paper problem separating unit 140 recognizes a vertical line (A) for distinguishing a paragraph from the test paper image, a margin (B) for distinguishing problems, and a problem number (C) that is the start of the problems. Test questions are generated by separating the images of the test questions.

Referring to FIGS. 2 and 3 again, in the text DB generation process (S400), the problem OCR unit 150 extracts the text of the test paper problem from the test paper problems stored in the test paper problem DB 115 to test text of the test paper problem. The test paper question-specific text DB 117 is stored in the DB 117.

Then, the problem similarity measurement unit 160 compares the test paper question text with the question text and measures similarity by performing the test paper problem measurement result DB generation process (S500). And measured similarities are stored in the measurement result DB (119) for each test paper problem as measurement result information for each test paper problem to generate a measurement result DB (119) for each test paper problem.

In addition, the problem likelihood measurer 160 receives a specific mathematical symbol, a mathematical formula, a word, a mathematical term, and at least one comparison target among a problem solving principle and a content from a user, and assigns the designated comparison target to priority. Problem similarity can be extracted. For example, you can designate mathematical symbols such as square root (√), integral (∫), sigma (∑), etc. as the first comparison targets (if the mathematical symbols in the extracted problem are different, you can specify only the symbols you want. ), A proportional expression, or the like, may be designated as the second comparison object, and mathematical terms such as "congruent triangle", "joint condition", "radius", and "debt" may be designated as the third comparison object. The principle and content of the problem can be specified as the fourth comparison object. The solution to the problem is, for example, asking whether the element inclusion symbol can be correctly represented by distinguishing between elements included in a given set and those not. Or a brief description of the problem, such as "choose elements in a given set." This allows the user to specify specific details and use the machine learning engine. Since the problem principle and the term are connected to each problem, it can be learned through machine learning how the problem principle and the term are formed. In addition, by providing test questions to machine learning engines that have learned the principles and terms associated with the extracted problems, they can find out the principles and terms in which the test questions appear, and provide these problems and variations of terms. 7 illustrates an example of a solution database associated with a problem, and FIG. 8 illustrates an example of a term database connected to a problem. Meanwhile, at least one of the above-described first to fourth comparison objects may be specified, and Priorities can also be specified. For example, only the first comparison target and the second comparison target can be designated, the user can specify contents desired within the designated comparison target, and the first comparison target should be designated first or the second comparison target preferentially. You can choose whether or not to specify it.

Thereafter, when a request for a search for a similar question or a question for a test question is input from a manager or a questionnaire, an interface screen for providing an interface that allows the measurement result inquiry and selection unit 180 to search for and output a similar test question is displayed. Outputs and retrieves and outputs the similarity and similarity correspondence problem measured for each test paper problem according to the user's command, and performs the measurement result inquiry and selection process (S600) of outputting the problem selected by the user as a question.

In this case, when the Internet service is provided, a search request for a similar problem for each test question or a test question request manager terminal (see 3, 6) or a questionnaire terminal (see 4, 6) may be input from an administrator or a questionnaire. . In addition, the questionnaire terminals 4 may include a teacher's terminal for preparing a test and a terminal for students who have taken a test or students who want to study a similar problem for each test paper question.

5 is a configuration diagram of the similar problem extraction server 1.

As shown in FIG. 5, the similar problem extracting server 1 is a control unit 10 as a central processing unit, a similar problem extracting apparatus 100 for providing an operation program executed by the control unit 10 and a similar problem extracting service 100. ) Is implemented by software storage unit 60 is stored, the input unit 30 configured to input data and user control commands, the display unit 40 for displaying the internal operation process and communication with the outside is required In this case, it may be configured as a server computer including a communication unit 50 to perform communication with the outside. Accordingly, the similar problem extracting apparatus 100 may be manufactured as a cross media in which codes executed by a computer are executed. Alternatively, the similar problem extracting apparatus 100 may be manufactured as a hardware apparatus to which an FPGA or the like is applied.

6 is a configuration diagram of a similar problem extraction service providing system for providing a similar problem extraction service through the Internet using the similar problem extraction server 1 equipped with the similar problem extraction apparatus 100.

As shown in FIG. 6, the similar problem extraction service providing system is configured such that the similar problem extraction server 1, the management company terminal 3, and the questionnaire terminal 4 communicate with each other through the communication network 9.

By such a configuration, the similar problem extracting server 1 may search for and study similar problems with the questions of the extracted test paper, or extract a similar problem as new test questions so that a student or teacher can make a test question. do.

Although the technical spirit of the present invention described above has been described in detail in a preferred embodiment, it should be noted that the above-described embodiment is for the purpose of description and not of limitation. In addition, those skilled in the art will understand that various embodiments are possible within the scope of the technical idea of the present invention. Therefore, the true technical protection scope of the present invention will be defined by the technical spirit of the appended claims.

100: Similar problem extraction device
1: Similar problem extraction server
3: manager terminal
4: questions terminal

Claims (5)

Test result by question information, which is similar information with similarity generated by comparing text information by question, test paper image, test question, test question problem text extracted from test question, text by question and text by question A test question DB unit 110 for storing the database;
A problem-specific text extraction unit 120 for extracting problem-specific text information and storing it in the test question DB unit 110;
An administrator scan unit 130 for storing the scanned test paper image in the test question DB unit 110;
A test paper problem separating unit 140 for processing the test paper image to separate test paper problems which are image information for each problem, and then storing the test paper images in the test question DB unit 110;
A problem OCR unit 150 for extracting text for each test paper question from the test paper questions and storing the test paper text in the test question DB unit 110;
A problem similarity measurer 160 which compares the text for each test paper question and the text information for each question, measures a problem similarity, extracts each compared problem as a similar problem, and stores the similar problem in the test question DB unit 110; And
And a measurement result inquiry and selection unit 170 for retrieving similar problems having similarity for each test paper question, and selecting and extracting similar problems to solve a question.
The problem OCR unit 150,
Designate mathematical symbols and mathematical formulas as main components and numbers as variables to extract text for each test questionnaire.
The problem similarity measuring unit 160,
A user is assigned a comparison target of at least one of a specific mathematical symbol, a mathematical formula, a mathematical term, and a solution principle and content of the problem, and extracts a problem similarity based on the specified comparison target as a priority;
The object of comparison of the problem solving principle and the content is to extract the problem similarity using the machine learning engine,
The test paper problem separator 140,
The test paper image is divided into vertical lines, margins, and question numbers, and the image is divided into test paper questions.
The problem likelihood measuring unit 160 includes a first comparison object including a mathematical symbol, a second comparison object including a mathematical structural formula, a third comparison object including a mathematical term, and a solution principle and content of the problem. In the case of designating a fourth comparison target, the priority between the first comparison target, the second comparison target, the third comparison target, and the fourth comparison target is specified,
Solving principles and contents of the problem included in the fourth comparison target is made of sentences,
The problem likelihood measurement unit 160, based on the pre-learned content about the principle and the solution of the problem through the machine learning engine to grasp the configuration of the terms configured in the fourth comparison target,
The machine learning engine is learned through machine learning how the solution principle and term are formed for each test question, and the solution principle and term are connected for each test question to identify the solution principle and term indicated by the test question. ,
The test question DB unit 110,
A problem-specific text DB 111 for storing the problem-specific text information extracted by the problem-specific text extraction unit 120;
A test paper DB 113 for storing the scanned test paper image output by the manager scanning unit 130;
A test paper problem DB 115 for storing the test paper problem, which is image information for each test generated by the test paper problem separating unit 140, by image-processing the test paper image for each test paper number;
A test paper question-specific text DB (117) for storing the test paper question-specific text extracted by the question OCR unit 150 from the test paper questions stored in the test paper question DB 115; And
The question similarity measurement unit 160 stores the similarity generated by comparing the test paper problem text and the text of problem paper with the test text problem test result DB 119 for storing the test text problem test result information. Is configured to include,
The test paper image stored in the test paper DB 113 is stored with school / year / grade / semester / middle or end information,
The problem-specific text information is extracted from the problems of the question bank and stored, and similar problems, including a question owned by the math problem DB unit of the question of the question bank.
delete delete delete Test question DB section 110, including a question-specific text DB (111), a test paper DB (113), a test paper question DB (115), a test paper question-specific text DB (117) and a test paper question-specific measurement results DB (119), Problem-specific text extraction unit 120, administrator scan unit 130, test paper problem separation unit 140, problem OCR unit 150, problem similarity measurement unit 160 and measurement results inquiry and selection unit 170 In the similar problem extraction method by the similar problem extraction apparatus 100,
A problem-specific dex DB generation process (S100) of storing the problem-specific text information extracted from the pre-stored and registered problems by the problem-specific text extraction unit 120 in the problem-specific text DB 111;
A test paper DB generation process of classifying the scanned test paper image by the manager scanning unit 130 and storing it in the test paper DB 113 (S200);
A test paper problem DB generation process of generating a test paper problem that is image information for each test by the test paper problem separating unit 140 by processing the image of the test paper and storing the test paper problem DB 115 for each test number per test paper;
A problem paper text generation process (S400) for the problem OCR unit 150 to extract the text for each test paper problem from the test paper questions stored in the test paper problem DB 115 and to store the text in the test paper problem text DB 117;
The problem likelihood measurer 160 compares the test paper text for each question and the text for each problem, and measures the problem similarity, and the test paper has the problem paper having the text for each problem and the measured problem similarity as the test paper problem measurement result information. A test sheet measurement result DB generation process stored in the problem measurement result DB 119 (S500); And
The measurement result inquiry and selection unit 180 searches for and outputs the similarity and similarity correspondence problem measured for each test paper problem according to the input command, and the measurement result inquiry for outputting a problem corresponding to the input selection signal as a question. Including the selection process (S600),
The text DB generation process for each test paper question (S400),
Through the problem OCR unit 150, the mathematical symbols and mathematical formulas are designated as main components and numbers are designated as variables to extract text for each test paper question,
The measurement result DB generation process for each test paper problem (S500),
The user is assigned a specific mathematical symbol, a mathematical formula, a word, a phrase expression pattern, and a comparison target of at least one of a problem solving principle and a content, and extracts a problem similarity based on the designated comparison target as a priority.
The object of comparison of the problem solving principle and the content is made by extracting the problem similarity using the machine learning engine,
The test paper problem extracted by the test paper problem separation unit 140 in the test paper problem DB generation process (S300) is a problem-specific image extracted by separating the test paper image based on vertical lines, margins, and problem numbers.
In the test paper DB measurement result DB generation process (S500), the first comparison target including a mathematical symbol, the second comparison target including a mathematical structural formula, the third comparison target including a mathematical term and the solution principle In the case of designating each of the fourth comparison targets including the contents, the priority of the first comparison target, the second comparison target, the third comparison target, and the fourth comparison target is designated.
Solving principles and contents of the problem included in the fourth comparison target is made of sentences,
In the test paper DB measurement result DB generation process (S500), based on the pre-learned content about the principle and the solution of the problem through the machine learning engine to identify the configuration of the terms configured in the fourth comparison target,
The machine learning engine is learned through machine learning how the solution principle and term are formed for each test question, and the solution principle and term are connected for each test question to identify the solution principle and term indicated by the test question. ,
The test question DB unit 110,
A problem-specific text DB 111 for storing problem-specific text information extracted by the problem-specific text extracting unit 120;
A test paper DB 113 for storing the scanned test paper image output from the manager scanning unit 130;
A test paper problem DB 115 for storing the test paper problem, which is image information for each test generated by the test paper problem separating unit 140, by image-processing the test paper image for each test paper number;
A test paper question-specific text DB 117 for storing a test paper question-specific text extracted from the test paper questions stored in the test paper question DB 115 by the problem OCR unit 150; And
The question similarity measurement unit 160 stores the similarity generated by comparing the test paper problem text and the text of problem paper with the test text problem test result DB 119 for storing the test text problem test result information. Is configured to include,
The test paper image stored in the test paper DB 113 is stored with school / year / grade / semester / middle or end information,
The problem-specific text information is extracted and stored in advance from the problems of the question bank, the similar problem extraction method, characterized in that it includes problems owned by the math problem DB unit.
KR1020170152320A 2016-12-22 2017-11-15 Apparatus and method for extracting similar test problem using recognition of test paper KR102004180B1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
KR1020160176621 2016-12-22
KR20160176621 2016-12-22

Publications (2)

Publication Number Publication Date
KR20180073440A KR20180073440A (en) 2018-07-02
KR102004180B1 true KR102004180B1 (en) 2019-10-01

Family

ID=62914221

Family Applications (1)

Application Number Title Priority Date Filing Date
KR1020170152320A KR102004180B1 (en) 2016-12-22 2017-11-15 Apparatus and method for extracting similar test problem using recognition of test paper

Country Status (1)

Country Link
KR (1) KR102004180B1 (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20220136592A (en) * 2021-03-31 2022-10-11 주식회사 잘라 A method for providing user-customized education contents

Families Citing this family (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109003492B (en) * 2018-07-25 2021-01-05 厦门大学附属心血管病医院(厦门市心脏中心) Topic selection device and terminal equipment
KR102048638B1 (en) * 2018-08-31 2019-11-25 망고슬래브 주식회사 Method and system for recognizing content
CN112989760A (en) * 2019-12-17 2021-06-18 北京一起教育信息咨询有限责任公司 Method and device for labeling subjects, storage medium and electronic equipment
US20210241644A1 (en) * 2020-02-03 2021-08-05 St Unitas Co., Ltd. Apparatus, method and recording medium storing command for supporting learning
CN112364679A (en) * 2020-09-04 2021-02-12 联想(北京)有限公司 Image area identification method and electronic equipment
CN113821712A (en) * 2021-09-24 2021-12-21 作业帮教育科技（北京）有限公司 Method for scanning and searching questions, intelligent scanning pen and storage medium
CN115454951B (en) * 2022-10-18 2023-06-30 读书郎教育科技有限公司 Teaching data sharing system

Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR101229860B1 (en) * 2011-10-20 2013-02-05 주식회사 매쓰홀릭 System and method to support e-learning

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7305129B2 (en) * 2003-01-29 2007-12-04 Microsoft Corporation Methods and apparatus for populating electronic forms from scanned documents
KR20040015786A (en) 2004-01-29 2004-02-19 박철서 Automatic setting questions for examination process in internet
KR20120048438A (en) * 2010-11-05 2012-05-15 (주)스마트쌤 System of method of creating paper with question identification information
KR20130089998A (en) * 2012-01-27 2013-08-13 김범수 System of providing study information and method of providing study information learner with study information using the same
KR20150025232A (en) * 2013-08-28 2015-03-10 인하대학교 산학협력단 Intelligent tutoring systems, methods, server and recording medium

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR101229860B1 (en) * 2011-10-20 2013-02-05 주식회사 매쓰홀릭 System and method to support e-learning

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20220136592A (en) * 2021-03-31 2022-10-11 주식회사 잘라 A method for providing user-customized education contents
KR102629686B1 (en) * 2021-03-31 2024-01-29 주식회사 고투무브 A method for providing user-customized education contents

Also Published As

Publication number Publication date
KR20180073440A (en) 2018-07-02

Similar Documents

Publication Publication Date Title
KR102004180B1 (en) Apparatus and method for extracting similar test problem using recognition of test paper
US11508251B2 (en) Method and system for intelligent identification and correction of questions
CN111144191B (en) Font identification method, font identification device, electronic equipment and storage medium
KR20190142287A (en) Method for recommending related problem based on meta data
CN104503998B (en) For the kind identification method and device of user query sentence
CN112035675A (en) Medical text labeling method, device, equipment and storage medium
JP7077483B2 (en) Problem correction methods, devices, electronic devices and storage media for mental arithmetic problems
KR101812763B1 (en) System for generating examination of korean spelling and analyzing test result thereof
CN111767382A (en) Method and device for generating feedback information and terminal equipment
CN112381099A (en) Question recording system based on digital education resources
KR20200031007A (en) Method building assessment infomation according to curriculum assessment and method providing thereof
CN112347997A (en) Test question detection and identification method and device, electronic equipment and medium
US10424217B1 (en) Systems and methods for ability-appropriate text generation
CN109408175B (en) Real-time interaction method and system in general high-performance deep learning calculation engine
KR102185733B1 (en) Server and method for automatically generating profile
CN111160606A (en) Test question difficulty prediction method and related device
CN114117222A (en) Resume document matching method and device, computing device and storage medium
US20210390251A1 (en) Automatic generation of form application
KR20170113619A (en) Semi-automated systems and methods for evaluating responses
WO2020175662A1 (en) Dictionary creating device, dictionary creating method, and dictionary creating program
KR101607128B1 (en) Providing related problem method about incorrect answer
JP5877775B2 (en) Content management apparatus, content management system, content management method, program, and storage medium
US11967253B2 (en) Semi-automated evaluation of long answer exams
JP2021117659A (en) Identifying device, identifying method, program, and data structure
CN113569112A (en) Tutoring strategy providing method, system, device and medium based on question

Legal Events

Date Code Title Description
A201 Request for examination
E902 Notification of reason for refusal
E701 Decision to grant or registration of patent right
GRNT Written decision to grant