WO2024048881A1

WO2024048881A1 - Learning system, and method for operating learning application

Info

Publication number: WO2024048881A1
Application number: PCT/KR2023/002650
Authority: WO
Inventors: 이충훈
Original assignee: 주식회사 애드아이랩
Priority date: 2022-08-31
Filing date: 2023-02-24
Publication date: 2024-03-07

Abstract

A learning system according to one disclosed embodiment comprises: at least one processor for executing an application and artificial intelligence trained through a server; and at least one display which receives an input of a user and displays same, and which outputs an operation of the application, wherein the processor receives a problem file uploaded by a user, separates a problem area and an option area from the problem file on the basis of the trained artificial intelligence, extracts text or a picture from the separated problem area and the separated option area, requests problem solution information about the problem file from the server on the basis of the extracted text or picture, and outputs the received problem solution information through the display on the basis of the input of the user.

Description

How learning systems and learning applications work

The present invention relates to a learning system and a method of operating a learning application. More specifically, the present invention relates to a learning system that recognizes various problem files input by a user, recognizes the problem area and the preliminary area, and provides various services based on this. It is about how systems and learning applications operate.

Recently, as the enthusiasm for studying various tests such as curriculum tests, college entrance exams, and civil service exams and certifications increases, interest in study methods is also increasing.

Previously, workbooks or workbooks that were provided in paper form are now being provided to learners in PDF or various data formats due to the advancement of information and communication. However, typical edutech companies provide workbooks and worksheets that the company itself creates or databases as applications, causing the inconvenience of having to install another application again in order for learners to solve or organize workbooks from other edutechs.

In addition, learners had the inconvenience of having to separately organize incorrect answers or information solved in multiple workbooks, and had to perform the organizing work themselves, so-called single volume, to reduce the huge amount of studying just before the exam.

According to one disclosed embodiment, an operation is performed to distinguish problem areas and preliminary areas from various types of problem files input by a user, recognize data extracted from the differentiated areas, and provide various services necessary for learners. It concerns how learning systems and learning applications operate.

The learning system according to the disclosed embodiment includes at least one processor executing artificial intelligence and applications learned from a server; and at least one display that receives and displays a user's input and outputs an operation of the application, wherein the processor receives a problem file uploaded by the user and, based on the learned artificial intelligence, Dividing a problem area and a problem area in a problem file, extracting text or extracting a picture from the divided problem area and the separated problem area, and extracting the problem file from the server based on the extracted text or picture. Requesting problem solving information and outputting the received problem solving information through the display based on the user's input.

The processor distinguishes between the problem area and the problem area by converting the received problem file into an image file, first recognizing the problem area through the artificial intelligence in the converted image file, and image excluding the problem area. It may include recognizing as the problem area.

The processor generates problem data based on the extracted text or the extracted picture, and the problem data includes page information including the problem area and the selection area in the problem file, extracted from the problem area, and It may include at least one of a question number, question content included in the problem area, the number of questions extracted from the question paper area, and question content included in the question paper area.

The problem solving information includes at least one of other users' incorrect answer rate, average solving time, and problem difficulty generated based on semantic search using Knowledge Gragh, and is based on the correct answer entered by the user. It can be updated in real time.

The process distinguishes a problem area and a question area where the user's solution is entered, compares the question area where the user's solution is entered with the correct answer included in the problem solution information, and determines whether the answer is correct based on the comparison result. can be decided.

The processor recognizes only the problem area and the question area containing the wrong answer based on the problem solving information, deletes areas other than the problem area and the question area containing the incorrect answer, and deletes the problem area and the question area containing the incorrect answer. Only the area can be rearranged and output on the display.

When rearranging only the problem area and the selection area containing the incorrect answer, the processor may delete the input value containing the user's solution and output only the problem area and the selection area containing the incorrect answer through the display. there is.

The processor may control the display to weakly display the sound range of areas other than the problem area and the selection area containing the user's solution, and to delete the input value containing the user's solution.

A method of operating an application of a user terminal that communicates with an external server according to another disclosed embodiment includes an artificial intelligence and a network that have learned how to distinguish problem areas and problem areas based on learning data labeling problem areas or problem areas. connected through; display problem files uploaded by the user; distinguishing problem areas and preliminary areas in the problem file based on the learned artificial intelligence; Extracting text and extracting pictures from the divided problem area and the divided paper area; requesting the problem solving information of the problem file from the server based on the extracted text and picture; and outputting the received problem solving information through the display.

Distinguishing a problem area and a problem area among the problem files includes converting the received problem file into an image file; In the converted image file, the line area is first recognized through the artificial intelligence; It may include recognizing the image excluding the image area as the problem area.

The problem solving information is generated by the artificial intelligence included in the server to include at least one of other users' incorrect answer rate, average solving time, and problem difficulty generated based on semantic search using Knowledge gragh, , the server updates the problem solving information in real time based on the solution input by the user.

distinguishing between a problem area and a preliminary area where the user's solution is input; Compare the answer number included in the problem solving information with the answer number of the question area in which the user's solution is entered; It may further include determining whether the answer is correct based on the comparison result.

Based on the problem solving information, recognize only the problem area and the preliminary area containing incorrect answers; Delete areas other than the question area and exam area containing the incorrect answer; It may further include rearranging only the problem area and the test area containing the incorrect answer.

distinguishing between a problem area and a preliminary area where the user's solution is input; displaying a user interface to guide editing of the problem file; It may include outputting only a selection area containing the user's solution and a problem area corresponding thereto based on the user input input through the user interface through the display.

deleting the user's solution only from the selection area containing the user's solution and the problem area corresponding thereto, and outputting the solution through the display; When the user's additional solution is input, only the problem area and the selection area that do not include the user's solution are rearranged and output through the display.

The learning system and learning application operation method according to the disclosed embodiment can distinguish problem areas and preliminary areas in various types of problem files, recognize data extracted from the distinguished areas, and provide various services necessary for learners. there is.

In addition, the learning system and learning application operation method according to another disclosed embodiment provides convenience of problem solving such as automatic scoring, creation of incorrect answer notes, and incorrect answer rate of other learners based on problem solving performed by the learner, and only problems including incorrect answers are provided. You can provide a service that allows learners to solve problems again by editing them separately.

1 is a diagram illustrating the configuration of a learning system according to an embodiment of the disclosure.

Figure 2 is a control block diagram for explaining the configuration of a user terminal.

Figure 3 is a flowchart for explaining a method of operating a learning application according to an updated embodiment.

Figures 4 to 8 show an example to specifically explain the operation method of Figure 3.

Figure 9 is a flowchart to explain in more detail the operation of an application for another embodiment.

Figure 10 is a flow chart to explain various services that can be provided to users using problem solving information.

FIGS. 11 and 12 are diagrams for specifically explaining the flow chart illustrated in FIG. 10 .

Figure 13 is a diagram of another embodiment of performing a single-copy service.

Like reference numerals refer to like elements throughout the specification. This specification does not describe all elements of the embodiments, and general content or overlapping content between the embodiments in the technical field to which the present invention pertains is omitted. The term 'unit, module, member, block' used in the specification may be implemented as software or hardware, and depending on the embodiment, a plurality of 'unit, module, member, block' may be implemented as a single component, or It is also possible for one 'part, module, member, or block' to include multiple components.

Throughout the specification, when a part is said to be “connected” to another part, this includes not only direct connection but also indirect connection, and indirect connection includes connection through a wireless communication network. do.

Additionally, when a part "includes" a certain component, this means that it may further include other components rather than excluding other components, unless specifically stated to the contrary.

Terms such as first and second are used to distinguish one component from another component, and the components are not limited by the above-mentioned terms.

Singular expressions include plural expressions unless the context clearly makes an exception.

The identification code for each step is used for convenience of explanation. The identification code does not explain the order of each step, and each step may be performed differently from the specified order unless a specific order is clearly stated in the context. there is.

Hereinafter, the operating principle and embodiments of the present invention will be described with reference to the attached drawings.

FIG. 1 is a diagram for explaining the configuration of a learning system according to an embodiment of the disclosure, and FIG. 2 is a control block diagram for explaining the configuration of a user terminal. To avoid redundant explanation, they are explained together below.

Referring first to FIG. 1, the disclosed learning system transmits and receives data through a user terminal 10 of a learner (hereinafter referred to as a user) and a server 20 through a network.

Specifically, the user terminal 10 receives the problem file uploaded by the user in the form of data, and communicates the received problem file and information necessary for services and operations according to the disclosed embodiment with the server 20 through the network.

The user terminal 10 can download an application containing the learned artificial intelligence from the server 20 and recognize the problem file uploaded by the user through the application. The user terminal 10 recognizes the problem file as a problem area and a selection area, extracts text and pictures, and provides them to the server 20. The application can provide various services to the user by receiving problem solving information from the server 20 and displaying it on the display of the user terminal 10.

In Figure 1, the user terminal 10 is expressed as a tablet PC, but it is not necessarily limited thereto, and may be implemented as a computer or portable terminal that can connect to the server 20 through a network. Here, the computer includes, for example, a laptop equipped with a web browser, a desktop, a laptop, a tablet PC, a slate PC, etc., and the portable terminal includes, for example, portability and mobility. Covered wireless communication devices include Personal Communication System (PCS), Global System for Mobile communications (GSM), Personal Digital Cellular (PDC), Personal Handyphone System (PHS), Personal Digital Assistant (PDA), and International Mobile Telecommunication (IMT). -2000, CDMA (Code Division Multiple Access)-2000, W-CDMA (W-Code Division Multiple Access), WiBro (Wireless Broadband Internet) terminal, smart phone, etc. based on all types of handhelds It may include wireless communication devices and wearable devices such as watches, rings, bracelets, anklets, necklaces, glasses, contact lenses, or head-mounted-device (HMD).

The server 20 is a configuration in which multiple CPUs are integrated and provided at a specific location designated by an application supplier. It is not limited to an HTTP server and has a hardware configuration that operates through a request from a client, that is, the user terminal 10. it means. The server software not only distinguishes the problem area and the preliminary area of the problem file through artificial intelligence, which will be described in detail below, and extracts text and pictures, but also generates problem solving information and then sends it to the user terminal 10. It may be composed of a series of programs provided, and may be continuously updated on the server 20. The CPUs constituting the server 20 do not necessarily need to be provided in the same space, and may be provided separately on a network to operate the server software.

Meanwhile, in this specification, the problem file input by the user is expressed as a problem collection, but the data received by the disclosed learning system is not limited to a file composed of problems for learners. Problem files contain various types of materials, such as worksheets containing information necessary for user learning, manuals and textbooks containing study guides, and these various types of materials are simply unified and explained as a problem file.

The problem solving information is generated by artificial intelligence provided in the server 20 based on semantic search using Knowledge gragh, and may specifically include at least one of the following: other users' incorrect answer rate, average solving time, and problem difficulty. You can. A detailed explanation of this will be provided later with reference to Figure 8.

Referring to FIG. 2, the user terminal 10 includes at least one processor 100 that performs artificial intelligence and applications learned through the server 20, learned artificial intelligence codes and problem-solving information, as well as the server 20 It includes a memory 110 that stores various data such as problem-solving information received from ) and at least one display 120 that receives and displays user input and outputs the operation of the application.

Specifically, the memory 110 includes non-volatile memory elements such as cache, read only memory (ROM), programmable ROM (PROM), erasable programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), and flash memory. Alternatively, it may be implemented as at least one of a volatile memory element such as RAM (Random Access Memory) or a storage medium such as a hard disk drive (HDD) or CD-ROM, but is not limited thereto. The memory 110 may be implemented as a separate chip from the processor 100, but is not necessarily limited to this, and may also be implemented as a single chip with the processor 100.

The display 120 includes a Digital Light Processing (DLP) panel, a Plasma Display Panel, a Liquid Crystal Display (LCD) panel, an Electro Luminescence (EL) panel, and an electrophoresis panel. It can be prepared as an Electrophoretic Display (EPD) panel, Electrochromic Display (ECD) panel, Light Emitting Diode (LED) panel, or Organic Light Emitting Diode (OLED) panel. It is not limited to this.

The display 120 may include a GUI (Graphical User Interface), that is, a software device, such as a touch pad, to receive user input. The touch pad is implemented as a touch screen panel (TSP) and can form a mutual layer structure with the display 120.

The processor 100 may be, for example, a central processing unit (CPU), and the components of the processor 100 described above may include software and/or a field programmable gate array (FPGA) and an application specific semiconductor (ASIC). Refers to hardware components such as Integrated Circuit.

In order to execute the operation of the application according to the disclosed embodiment, the processor 100 receives the problem file uploaded by the user, and the file receiving unit 101 temporarily stores the problem area in the problem file through learned artificial intelligence. After distinguishing the and line areas, the extraction unit 102 extracts text and/or pictures, generates problem data based on the extracted text and/or pictures, and transmits the generated problem data to the server 20 to provide user information. An information generation unit 103 that receives problem-solving information to be provided to the user and a so-called single-issue service that removes problem areas and preliminary paper areas that the user does not want and rearranges the remaining problem areas and preliminary paper areas based on the user's input. It may be divided into a file creation unit 104.

Specifically, the file receiving unit 101 may receive the problem file input by the user as data such as PDF, or may receive the problem file in the form of an image by being photographed by a camera provided in the user terminal 10.

The processor 100 stores problem files uploaded in various formats in the memory 110. In the processor 100, the extraction unit 102 divides the problem file into a problem area and a preliminary area through object recognition.

Specifically, the extraction unit 102 converts the problem file input in PDF format into image format. If a file has already been received in image form by taking a picture by the user, the process of converting it to image form can be omitted.

In the converted image file, the extraction unit 102 recognizes the problem area and the selection area through artificial intelligence previously learned through the server 20. In the disclosed embodiment, the artificial intelligence may be a YoLO (You only Look Once) neural network. Specifically, YoLO is a neural network specialized in object recognition that identifies objects in images or videos through deep learning, and recognizes objects through bounding boxes. YoLO is a type of CNN (Convolutional Neural Network) that can be trained by continuously repeating the process of going through the Conv Layer and Max Pooling Layer. However, according to the disclosed embodiment, the neural network that recognizes the problem area and the predetermined area is not necessarily limited to YoLO, and may include various artificial neural networks that can be used in object recognition, such as CNN and Fast YoLO.

Meanwhile, artificial intelligence according to the disclosed embodiment can learn how to distinguish between problem areas and preliminary areas through numerous learning data in the server 20. Specifically, artificial intelligence is learned through labeled learning data that bounds the preliminary area and the problem area in the problem file and determines whether the bounded area is the problem area or the preliminary area.

The learned artificial intelligence is transmitted to the user terminal 10 or stored in the server 20, and the extraction unit 102 uses the learned artificial intelligence to recognize objects for the problem file input from the user. In the disclosed embodiment, the extraction unit 102 first divides the paper area for each page of the problem file to create a bounding box, and then recognizes the remaining area excluding the paper area and the entire paper area as the problem area. A specific embodiment in which the extraction unit 102 recognizes the problem area and the problem area will be described later with reference to FIG. 6, etc.

When the problem file is divided into a problem area and a paper area for each page, the extraction unit 102 extracts text or pictures included in each area. For example, from the problem area, the extraction unit 102 can extract the problem number, problem content, formula, or a picture, graph, or view needed to solve the problem. In the selection area, numbers representing candidates and text (including numbers) containing correct and incorrect answers can be extracted.

The information generation unit 103 generates problem data based on the text extracted by the extraction unit 102.

Specifically, problem data refers to a variety of information that can be extracted from the recognized problem area and preliminary area, for example, page information, problem number extracted from the problem area, problem content included in the problem area, and extracted from the preliminary area. It may include at least one of the number of slices and the contents of the slices included in the slice area. A detailed description of data that may be included in the problem data will be described later with reference to FIG. 7, etc.

When problem data is generated, the information generation unit 103 transmits the problem data to the server 20. The server 20 generates problem solving information based on the received problem data.

Here, the problem solving information includes not only information about which test the problem file corresponds to, but also the correct answer to the problem recognized in the problem file, the average solving time and error rate of other users for problems recognized as problem areas, and Including problem difficulty level, etc.

The server 20 according to one embodiment may use semantic search using Knowledge Gragh to generate problem solving information. Semantic search means that search queries search for keywords in problem data through learned artificial intelligence, identify the intent of the problem data, and search for problem-solving information in problem files stored in advance or located on the network. As an example, the disclosed server 20 finds an object using the subject and predicate included in the text extracted from the problem data, then converts the object of the ontology into a triple subject and continuously searches the linked data. Through this, the server 20 searches for data that clearly matches the problem file based on the problem data and the correct answer information matched thereto, and then generates problem solving information based on the correct answer information.

Meanwhile, the disclosed embodiment does not necessarily need to use semantic search using Knowledge Gragh to generate problem-solving information. The server 20 may use various types of artificial intelligence to generate problem-solving information, or it may use problem file information that has already been converted into data from when the problem file was created. In other words, it is sufficient for the disclosed server 20 to process data to provide necessary services to users through problem-solving information, and it is not necessary to use only semantic search.

The problem solving information may be information including whether the problem file is related to a certain test.

Additionally, the problem solving information may also include page information of the problem file, problem number, and information about the correct answer to the problem area recognized through the selection. If the problem data includes the user's solution information, the problem solution information may also include whether the user's input value is a correct answer or an incorrect answer.

The problem solving information may include the error rate or difficulty level of the problem area based on other users' incorrect answers that are stored and updated in advance, and may also include information about the average solving time of other users.

The server 20 transmits problem-solving information containing various data to the user terminal 10, and the information generator 103 can visualize the problem-solving information and provide it to the user through the display 120. An embodiment of providing problem-solving information will be described later with reference to FIG. 8, etc.

The file creation unit 104 can create a new file by editing the problem file in various ways based on user input.

Specifically, the user may want to extract only the problem areas and areas where the user made a mistake and create his or her own incorrect answer summary note. In order to meet these user needs, the file creation unit 104 may utilize pre-extracted problem areas and preset areas to create new problem files that meet the needs of individual users.

As an example, the file creation unit 104 extracts only the problem area and the test area in which an incorrect answer was entered based on the problem solving information, and deletes the problem area and the test paper area in which the correct answer was entered. The file creation unit 104 may rearrange the remaining problem areas and selected areas and then output them on the display 120.

As another example, the file creation unit 104 may delete and rearrange only the areas marked by the user in the problem area and the paper area. As another example, the area other than the problem area and the paper area in which the handwriting was entered in the problem file may be weakened. You can also process it and create a new problem file.

The various services that the file creation unit 104 creates using problem solving information will be described later with reference to FIG. 11 and the like.

Meanwhile, the configuration of the processor 100 described in FIG. 2, that is, the file receiving unit 101, the extracting unit 102, the information generating unit 103, and the file generating unit 104, is software for specifically explaining the operation of the application. This is a general explanation, and is not necessarily distinguished through a separate hardware chip.

Figure 3 is a flowchart for explaining a method of operating a learning application according to an updated embodiment. Figures 4 to 8 show an example to specifically explain the operation method of Figure 3. To avoid redundant explanation, they are explained together below.

Referring first to Figure 3, the application receives a problem file to be uploaded by the user (200).

The application may output a user interface through the display 20 as shown in FIG. 4 so that the user can check the problem file uploaded. Referring to FIG. 4, the application may provide an icon 11 that allows the user to select whether the question file is included in a certain exam year or related to a certain exam, and as shown in FIG. A function to input a file name (12) can be provided.

As described above in Figure 2, the disclosed learning system generates problem data. While generating problem data, the disclosed learning system may refer to data corresponding to file names or icons classified by the user. In other words, in addition to information recognized through text in the problem file, problem data can also be generated based on the type of problem entered by the user.

When a user uploads a problem file, the application divides the problem file into a problem area and a preliminary area (201).

Specifically, the application converts the problem file into an image, first classifies the line area within the image through learned artificial intelligence, and then can classify the problem area.

Specifically, as shown in Figure 6, the application can recognize the part of the problem file related to 'Moon 1' as four preliminary areas and one problem area. The four test areas are divided into bounding boxes for each line, and the entire area other than the four test areas can be divided into a problem area in the form of a bounding box.

Meanwhile, the application converts the problem file into an image and displays a message saying "The uploaded problem is being recognized" to make the user aware that artificial intelligence is performing an operation to classify the problem area (120). It can be printed to . That is, image files divided by bounding boxes as shown in FIG. 6 are not output through the display 120 and may not be displayed to the user.

Once the problem area and the preliminary area are distinguished, the application extracts text and pictures from each area (203).

In the example of Figure 6, the application can extract "question 1" and the content of the problem as text in the problem area, and the question area can extract the content of each question as text, as well as the question number such as "①②③④". You can.

Unlike the example shown in FIG. 6, etc., if the problem file includes a graph or picture, the extracted graph or picture is recognized as a problem area and extracted.

Extracted text or pictures are converted into problem data. As the example shown in FIG. 7, the application may generate problem data as shown in FIG. 7. However, the generated problem data may not be visually displayed to the user.

Specifically, the application determines that the uploaded problem file is 'first page', the pixels occupied by the problem area correspond to [22, 740, 150, 125], the problem number is '1', and no views are included. You can create problem data with 4 choices. Additionally, the problem data may include the content of the problem and the content of each selection based on the extracted text.

The application transmits the generated problem data to the server 20 (204).

The server 20 generates problem solving information including at least one of other users' incorrect answer rate, average solving time, and problem difficulty generated based on semantic search using Knowledge gragh.

The generated problem-solving information is transmitted back to the application, and the application obtains the problem-solving information (205).

As in the example shown in Figure 8, the problem solving information includes information about the test type in which the problem related to the problem file is '2019 Civil Service Written Test', information that the user's input value is an 'incorrect answer', and information that other learners have solved the same problem. It includes information on the results of entering incorrect answers (30% incorrect answer rate), the average time required to solve the recognized problem (15 seconds), and the level of difficulty (level C) based on the incorrect answer rate and average time.

In Figure 8, it is assumed that the problem area and the preliminary paper area where the user inputs the solution to question 1 are transmitted. The server 20 may compare the correct answer of the problem solving information based on the user's input value and generate problem solving information including that the user's input value is an incorrect answer.

The application may output problem solving information through the display 120 (206).

There can be various ways to display problem-solving information. The application may display problem-solving information in simple text form as shown in FIG. 8, but may also provide problem-solving information to users using a visual user interface through various graphs, images, and avatars.

Meanwhile, the drawings shown in FIGS. 4 to 8 are merely examples for explaining the operation of the application, and the problem solving information may include various information that can be generated by recognizing the problem area and the selection area.

Figure 9 is a flowchart for specifically explaining the operation of an application according to another embodiment.

Before explaining the embodiment of FIG. 9, the flowchart of FIG. 9 assumes that the user terminal 10 has already distinguished the problem area and the preliminary area of the problem file and then obtained problem solving information from the server 20. . In addition, this is a flow chart for an embodiment in which the user terminal 10 obtains problem solving information that does not include incorrect answers from the server 20 and then receives the user's input.

Referring to Figure 9, the application receives user input (210).

Here, the user's input may mean a solution entered from a problem file uploaded by the user or an input value that the user checks as an important problem.

The application compares the user's input with previously obtained problem solving information (220).

The problem-solving information obtained in advance includes preliminary information corresponding to the correct answer, and the application can determine whether the user's input is the correct answer by re-recognizing the preliminary exam area containing the user's input value among the divided answer areas. .

Based on the comparison results, the application determines whether the answer is correct (231).

If the paper in the paper area containing the user's handwriting matches the correct answer included in the problem solving information, the application may display a screen indicating that the user's input is the correct answer. If the answer is incorrect, the application may display a screen indicating that the user's input is an incorrect answer.

Based on the user's input, the application updates problem solving information, including whether the answer is correct (232).

Updating problem solving information may be performed on the server 20. Since the problem solving information includes information about the incorrect answer rate of other users, the incorrect answer rate may change depending on whether the user's solution is correct or incorrect. The application transmits information about the correct or incorrect answer to the server 20 according to the user's solution. Problem solving information updated in the server 20 can be used when another user requests problem solving information for the same problem file.

Figure 10 is a flow chart to explain various services that can be provided to users using problem solving information. FIGS. 11 and 12 are diagrams for specifically explaining the flow chart illustrated in FIG. 10 .

Referring first to FIG. 10, the application receives a user input for editing a problem file (300).

As shown in Figure 12, user input leading to editing of a problem file can be received by the application through various user interfaces.

When a command for editing a problem file is received, the application recognizes the problem area and the preliminary area containing the user solution (310).

Here, the user solution includes various forms input by the user. As shown in the second screen 122 of FIG. 11, the user can input a user solution into the problem area and question paper area of Problem 1 through a user interface such as a fluorescent color display. The application can recognize the problem area and the preliminary area that contain the user's solution.

A service that weakly expresses shading in areas not included in the user's solution can be provided (322).

For example, the third screen 123 of FIG. 11 weakly displays an area that does not include the user's solution. That is, the first screen 121 in FIG. 11 is a screen in which the uploaded problem file is output through the display 120, the second screen 122 is a problem file in which the user solution is input, and the third screen 123 is a screen in which the uploaded problem file is output through the display 120. This is a screen that provides a service that displays weak shading in areas other than the problem area where the user's solution is entered and the selection area.

The application can output with the user solution deleted (343).

Specifically, while outputting the third screen 123 of FIG. 11, the application provides a service that allows the user to solve the problem again by outputting only the problem area and the selection area in which the user solution entered in the second screen 122 has been deleted. can do.

Referring again to FIG. 10, the application may perform a service of deleting a problem and an optional area for which the user has to solve (321) and rearranging the problem area and an optional area (330).

As shown in the first screen 124 of FIG. 12, the application can recognize five problem areas and question areas recognized as incorrect answers during the user's solution. For example, the application can compare the problem solving information with the user's solution and determine that five problems have incorrect answers.

As shown in the second screen 125 of FIG. 12, the application can create a new file that preserves only the problem areas and selected areas recognized as incorrect answers. Here, the application can rearrange the problem area and preset area to be preserved by considering the size of the problem file and the size of each area.

Relocation of the problem area and preset area is performed based on problem data generated when the application recognizes the problem file, and relocation is performed according to page information, size of the area, and size of the entire page. Relocation may be performed by leaving an area on the right as shown in the second screen 125 of FIG. 12, but this rearrangement is only an example and rearrangement may be performed in various ways.

When relocation is completed, the application outputs the relocated file with the user's solution remaining as is (341), and deletes and outputs the user's solution as shown in the second screen 125 of FIG. 12 (342).

As an example, the application may include a function to re-edit an edited problem file for a user who studies the same problem file multiple times.

As shown in the first screen 126 of FIG. 13, the user can edit the problem file after inputting user solutions into several problem areas and selection areas of the problem included in the uploaded problem file.

Assuming that the user entered a user solution (processed with a highlighter as shown in FIG. 13) for

problems

1, 2, 7, and 8, the application displays a solution area containing the user solution, as shown in the second screen 127 of FIG. 13. You can leave only the and corresponding problem areas, and delete the remaining problem areas and preset areas. That is, after the application recognizes the problem area and the paper area in which the user's solution is input in the problem file, it can rearrange only the problem and paper area in the area containing the user's solution and output it as shown in the second screen 127 of FIG. 13. there is.

Afterwards, the user can again input another user's solution into the relocated problem file (second screen 127 in FIG. 13). At this time, the application can edit the problem file while leaving the problem area and optional area that do not include the user's solution.

As shown in the third screen 128 of FIG. 13, the application outputs a rearranged edit file by deleting the problem area and the preliminary paper area containing the user solution and preserving only the problem area and the preliminary paper area that do not contain the user solution. You can.

However, in order for the application to perform different editing operations, such as the second 127 in Figure 13 or the third 128 in Figure 13, a user input command is required, and various editing services are provided according to the user's input command. It can be done.

The disclosed learning system and the application operating through the learning system can use recognized problem areas and preliminary areas and perform various functions that provide convenience to learners through solutions that can be entered by users in various ways, and through this, users can increase learning efficiency.

Claims

At least one processor executing artificial intelligence and applications learned from the server; and

At least one display that receives and displays user input and outputs operations of the application,

The processor,

Receiving the problem file uploaded by the user,

Based on the learned artificial intelligence, distinguish problem areas and preliminary areas among the problem files,

Extracting text or extracting pictures from the divided problem area and the divided paper area,

Requesting the problem solving information of the problem file from the server based on the extracted text or picture,

A learning system comprising: outputting the received problem-solving information through the display based on the user's input.
According to clause 1,

The processor distinguishes between the problem area and the problem area by:

Converting the received problem file into an image file,

In the converted image file, the line area is first recognized through the artificial intelligence,

A learning system comprising: recognizing an image excluding the image area as the problem area.
According to clause 1,

The processor,

Generating problem data based on the extracted text or the extracted picture,

The problem data is,

Page information including the problem area and the selection area in the problem file,

Problem number extracted from the above problem area,

Problem content included in the above problem areas;

The number of blood extracts extracted from the blood fat area,

A learning system including at least one of the preliminary content included in the preliminary information area.
According to clause 1,

The above problem-solving information is,

Contains at least one of other users' incorrect answer rate, average solution time, and problem difficulty generated based on semantic search using Knowledge gragh,

A learning system that is updated in real time based on the correct answer entered by the user.
According to clause 1,

The process is,

Distinguish between a problem area and a preliminary area where the user's solution is input,

Compare the answer area included in the problem solving information with the answer area where the user's solution is entered,

A learning system that determines whether the correct answer is based on the comparison results.
According to clause 5,

The processor,

Based on the problem solving information, recognize only the problem area and the preliminary area containing incorrect answers,

A learning system that deletes areas other than the problem area and the selected area containing the incorrect answer, rearranges only the problem area and the selected area containing the incorrect answer, and outputs them on the display.
According to clause 6,

The processor,

A learning system that deletes the input value containing the user's solution and outputs only the problem area and the test area containing the incorrect answer through the display when rearranging only the problem area and the test area containing the incorrect answer.
According to clause 5,

The processor,

Weakly display the sound range of areas other than the problem area and the preliminary area containing the user's solution,

A learning system that controls the display to delete input values containing the user's solution.
In a method of operating an application of a user terminal that communicates with an external server,

It is connected through a network with artificial intelligence that has learned how to distinguish problem areas and predicted areas based on learning data labeling problem areas or predicted areas;

display problem files uploaded by the user;

distinguishing problem areas and preliminary areas in the problem file based on the learned artificial intelligence;

Extracting text and extracting pictures from the divided problem area and the divided paper area;

requesting the problem solving information of the problem file from the server based on the extracted text and picture;

An application operation method comprising: outputting the received problem-solving information through the display.
According to clause 9,

Distinguishing between problem areas and preliminary areas among the problem files is as follows:

converting the received problem file into an image file;

In the converted image file, the line area is first recognized through the artificial intelligence;

An application operation method comprising: recognizing an image excluding the image area as the problem area.
According to clause 9,

The above problem-solving information is,

The artificial intelligence included in the server generates information to include at least one of other users' error rate, average solution time, and problem difficulty generated based on semantic search using Knowledge gragh,

The server is

An application operation method comprising: updating the problem solving information in real time based on the solution input by the user.
According to clause 9,

distinguishing between a problem area and a preliminary area where the user's solution is input;

Compare the answer number included in the problem solving information with the answer number of the question area in which the user's solution is entered;

A method of operating an application further comprising: determining whether the correct answer is based on the comparison result.
According to clause 12,

Based on the problem solving information, recognize only the problem area and the preliminary area containing incorrect answers;

Delete areas other than the question area and exam area containing the incorrect answer;

A method of operating an application further comprising rearranging only the problem area and the exam area containing the incorrect answer.
According to clause 9,

distinguishing between a problem area and a preliminary area where the user's solution is input;

displaying a user interface to guide editing of the problem file;

An application operation method comprising: outputting only a selection area containing the user's solution and a problem area corresponding thereto through the display based on user input input through the user interface.
According to clause 14,

deleting the user's solution only from the selection area containing the user's solution and the problem area corresponding thereto, and outputting the solution through the display;

When the user's additional solution is input, only the problem area and the selection area that do not include the user's solution are rearranged and output through the display.