KR20190080317A - Method and apparatus for explanation of text and question - Google Patents

Abstract

The present invention relates to a method for explaining a text and a question which provides key information of a text for increasing text reading abilities of a user and a device thereof. The method for explaining a text and a question comprises the steps of: allowing an online education providing device to provide a text explanation to a user device; and allowing the online education providing device to provide a question explanation to the user device. The text explanation can be performed based on a keyword and a key sentence of the text analyzed by a sentence structure analysis device.

Description

TECHNICAL FIELD [0001] The present invention relates to a method and apparatus for fingerprinting,

FIELD OF THE INVENTION The present invention relates to a fingerprint and problem explaining method and apparatus. And more particularly, to a method and apparatus for a user to more easily acquire information related to a fingerprint and a problem through analysis of a fingerprint and analysis of a problem.

Machine translation has been recognized as one of the most difficult tasks in natural language processing and artificial intelligence. This is because accurate translation is impossible without understanding the text. For that reason, researchers once reached a negative conclusion about machine translation.

Various methods have been studied for machine translation so far. These studies mainly focus on translation methods such as vocabulary of two languages, dictionary of phrases, dictionary of idioms, grammar of individual language or translation rules and translation dictionary, Various types of knowledge and algorithms have been applied for different purposes.

As deep learning has been applied recently, data-based artificial intelligence, which is not dependent on people, is transformed into a way to learn and understand language by oneself. Google has built a language model based on the accumulated data from web services and has used about 100 billion words from textual information obtained through news services for years in machine learning.

A variety of techniques have been introduced to help users learn based on their understanding of these languages.

It is an object of the present invention to solve all the problems described above.

Another object of the present invention is to provide fingerprint and problem information in a more stereoscopic manner in order to facilitate understanding of the fingerprint and the problem of the user.

Another object of the present invention is to provide the user with key information of the fingerprint in order to improve the reading ability of the fingerprint.

In order to accomplish the above object, a representative structure of the present invention is as follows.

According to an embodiment of the present invention, the fingerprint and problem explaining method may include the step of providing the fingerprint explanation to the user device by the online education providing device and the step of providing the problem explanation to the user device by the online education providing device The fingerprint commentary may be performed based on the fingerprint keyword and key sentence analyzed by the parsing device.

According to another embodiment of the present invention, an on-line training provision apparatus for performing fingerprint and problem interpretation includes a communication unit for communication with a user apparatus and a processor operatively connected to the communication unit, The fingerprint commentary is provided to the user device, and the fingerprint commentary is implemented based on a key word and a key sentence of the fingerprint analyzed by the parsing device.

According to the present invention, information on fingerprints and problems can be provided more stereoscopically to help understand fingerprints and problems.

In addition, according to the present invention, key information of a fingerprint can be provided for a user to improve the reading ability of the fingerprint.

1 is a conceptual diagram illustrating an online learning providing system according to an embodiment of the present invention.
2 is a conceptual diagram illustrating a method of providing key sentence information and keyword information on a fingerprint according to an embodiment of the present invention.
3 is a conceptual diagram illustrating a method of providing fingerprint information related to a problem according to an embodiment of the present invention.
4 is a conceptual diagram illustrating a keyword extraction method according to an embodiment of the present invention.
5 is a conceptual diagram illustrating a key sentence extraction method according to an embodiment of the present invention.
6 is a conceptual diagram illustrating a method for providing keyword information and key sentence information according to an embodiment of the present invention.

The following detailed description of the invention refers to the accompanying drawings, which illustrate, by way of illustration, specific embodiments in which the invention may be practiced. These embodiments are described in sufficient detail to enable those skilled in the art to practice the invention. It should be understood that the various embodiments of the present invention are different, but need not be mutually exclusive. For example, the specific shapes, structures, and characteristics described herein may be implemented by changing from one embodiment to another without departing from the spirit and scope of the invention. It should also be understood that the location or arrangement of individual components within each embodiment may be varied without departing from the spirit and scope of the present invention. Therefore, the following detailed description is not to be taken in a limiting sense, and the scope of the present invention should be construed as encompassing the scope of the appended claims and all equivalents thereof. In the drawings, like reference numbers designate the same or similar components throughout the several views.

Hereinafter, various embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those skilled in the art can easily carry out the present invention.

1 is a conceptual diagram illustrating an online learning providing system according to an embodiment of the present invention.

1, an online learning providing system for providing online training to a user is disclosed.

Referring to FIG. 1, the online learning providing system may include an online education providing apparatus 100, a parsing apparatus 120, and a user apparatus 140.

The online education provision apparatus 100 may be implemented to store problem information according to a fingerprint and a fingerprint, and to provide problem information according to the fingerprint and the fingerprint to the user apparatus. The online education provision apparatus 100 can receive problem information according to fingerprints and fingerprints learned by the syntax learning apparatus 120 and transmit problem information according to a fingerprint and a fingerprint to the user apparatus.

Also, the online education provision apparatus 100 can provide various education functions to the user apparatus. For example, when a specific button is clicked, explanation of a fingerprint starts, and information about a keyword and a key sentence of the fingerprint can be provided.

Also, the online education provision apparatus 100 can provide commentary information related to the problem to the user apparatus. For example, there may be a particular option in a problem associated with a fingerprint, and a fingerprint-related portion associated with a particular option may be provided with a separate indication.

The user device 140 may be implemented to receive problem information according to a fingerprint and a fingerprint provided by the online education providing apparatus 100 and to provide information for learning. As described above, information on keywords and key sentences of the fingerprint can be provided on the screen when the user's fingerprint is read. Information about keywords and key senses can be provided based on a separate display such as a highlight display.

In addition, when a specific button is clicked on the user device 140, explanation of the fingerprint is started, and information on the fingerprint can be provided by setting the speed of the screen change for comment on the fingerprint and whether or not it is repeated.

In addition to this, the user device 140 can provide a description of the problem together with a sentence of a fingerprint associated with the problem in providing the problem description information. For example, information about the sentence of the fingerprint that is relevant to the choice of the particular problem may be provided on the user device 140. [

The parsing device 120 may be implemented to generate additional information for learning based on learning about a fingerprint and a fingerprint-related problem. The parsing apparatus 120 can extract keywords and key sentences based on learning about fingerprints. The parsing apparatus 120 can classify fingerprints into sentence units, word units, morpheme units, and the like.

In addition, the parsing apparatus 120 can analyze each sentence and determine the relationship between the sentences. For example, the relationship between sentences can be determined based on sentence semantic information, inter-sentence conjunction information, and the like. For example, the sentence semantic information is obtained based on the words constituting the sentence, and the inter-sentence conjunction information can be obtained based on the information about the conjunction that connects the sentence and the sentence. The parsing apparatus 120 can extract the key sentence in the fingerprint based on the relationship between the sentences.

In addition, the parsing apparatus 120 can extract keywords based on the information about the relation between the sentences and the word information in the sentence. Information on the keyword and key sentence extracted by the parsing apparatus 120 may be transmitted to the online teaching provision apparatus 100 and provided for learning by the user.

In addition, the parsing device 120 can extract sentences associated with the problem on the fingerprint based on an analysis of the problem. For example, when sentences 1 to 10 constitute fingerprints and a sentence related to sentence 2 is provided as option 3 of problem 1, when proceeding with explanation of problem 2, sentence 2 is related to option 3, Display (e. G., Highlighting). ≪ / RTI >

2 is a conceptual diagram illustrating a method of providing key sentence information and keyword information on a fingerprint according to an embodiment of the present invention.

2, a method for providing key sentence information 200 and keyword information 250 on a fingerprint is disclosed.

Referring to FIG. 2, after the user solves the problem, commentary information on the fingerprint can be provided. Explanation of the fingerprint When providing the information, the user can start commenting on the fingerprint by clicking a specific button. The display speed and repeatability of the fingerprint can be determined according to the setting of the user.

The key sense information 200 and the keyword information 250 on the fingerprint are displayed on the screen while being emphasized in the order in which the user needs to find the key sentence information 200 and the keyword information 250 when reading the fingerprint .

By providing the key sentence information 200 and the keyword information 250 of the fingerprint, the user can quickly grasp the core contents of the entire fingerprint and can help improve the reading ability of the user.

3 is a conceptual diagram illustrating a method of providing fingerprint information related to a problem according to an embodiment of the present invention.

FIG. 3 discloses a method for providing fingerprint information related to a problem when explaining a problem is proceeded.

Referring to FIG. 3, when an explanation on a problem progresses, information 310, 320, and 330 regarding related fingerprints, which are fingerprints related to a problem option, may be provided with a separate indication. For example, if the problem is to pick the right information on the fingerprint, and the choice of the question "Suppose that General Admiral Yi won a great victory by burning over 200 enemy ships, . In this case, the relevant fingerprint information (310, 320, 330) related to 'Admiral Yi won a great victory by burning about 200 ships in battle with more than 500 fighters' Display). ≪ / RTI >

In this way, the user can quickly identify the part of the fingerprint corresponding to the option when providing a solution to the problem.

4 is a conceptual diagram illustrating a keyword extraction method according to an embodiment of the present invention.

In Fig. 4, a method of extracting a fingerprint keyword is disclosed.

Referring to FIG. 4, the fingerprint keyword may be extracted based on the repetition of the word in the fingerprint. The parsing apparatus 120 can extract repeated words on the fingerprint and count the number of repeated words over a threshold number of times.

The word A, the word B, the word C, and the word D may be determined as the primary keyword 410 of the fingerprint when the word A, the word B, the word C, and the word D are repeated more than the first threshold number of times in the fingerprint. The word E, the word F, the word G, and the word H can be determined as the secondary keyword 420 of the fingerprint when the word E, the word F, the word G, and the word H are repeated more than the second threshold number of times in the fingerprint.

The feedback for the learning of the primary keyword 410 and the secondary keyword 420 of the fingerprint can be performed. In order to learn about the actual keyword, the keyword trainer reads the fingerprint and selects a keyword that is considered as the keyword of the actual fingerprint among the primary keywords 410 as the selected keyword. If there is no keyword among the primary keywords 410 that the fingerprint trainer reads and reads the fingerprint, the keyword trainer can select the keyword that is considered as the keyword of the actual fingerprint among the secondary keywords 420 as the selection keyword 430. If there is no keyword in the secondary keyword 420 that the trainer reads and reads the fingerprint, the keyword trainer can directly select the keyword that is considered as the keyword on the fingerprint as the selection keyword 430.

The selected keyword 430 determined as the keyword analysis result of the keyword trainers can be transmitted to the parser again. Based on the selection keyword 430 determined as a result of the keyword analysis of the keyword trainers, the parsing device can learn the keyword. The parsing apparatus extracts the keyword from the fingerprint based on the number of repetitions of the selected keyword 430, the sentence including the selected keyword 430, the relation between the sentence including the selected keyword 430 and the surrounding sentence, the sentence weight, You can proceed to learn about the requirements for Learning about such keyword extraction can be performed continuously and in a large amount by generating a data set.

The understanding of the sentence meaning including the selection keyword 430 is based on the search of the dictionary database for the meaning of each word constituting the grammar structure (subject, predicate) for the sentence including the selection keyword 430 Lt; / RTI >

The relation between the sentence including the selection keyword 430 and the surrounding sentence can be determined in consideration of the understanding of the sentence meaning and the connection structure between sentences. For example, the semantics of a sentence and grammatical structure (for example, conjunctions) can be considered to determine whether a sentence relationship has a parallel relationship or a separate dependency. In the case of dependencies, the nature of specific dependencies, such as whether the relationship between claims and evidence can be determined.

The sentence weights can be determined in the fingerprint by considering the relationship between the sentences. In the case of dependency, the weights of the main sentence and subordinate sentences can be set differently.

This parsing apparatus is based on the number of repetitions of the selection keyword 430, the sentence including the selected keyword 430, the relation between the sentence including the selected keyword 430 and the surrounding sentence, The final keyword 440 can be extracted considering the understanding of the sentence in the fingerprint, the relationship between the sentences in the fingerprint, and the number of words. Based on the learning result, the final keyword 440 in the future fingerprint can be determined without the intermediate feedback of the separate keyword trainer.

For additional learning, keyword selection information of the user who learns fingerprints can also be added. If it is thought that the information is selected as a keyword on the fingerprint and is not an actual keyword, the learning users can input an additional selection keyword 430, and the user can input additional optional keyword 430 The keyword of the fingerprint can be determined.

The parsing device can extract keywords from the fingerprint more accurately based on the results of the first learning and the second learning on these keywords.

According to the embodiment of the present invention, the syntax analyzing apparatus can extend the range of the keyword according to the setting of the user. Keywords existing in the fingerprint can be prioritized according to importance. As described above, when the final keyword is determined in consideration of the understanding of the sentence in the fingerprint, the relation between the sentences in the fingerprint, the sentence weight, and the number of repetitions of the word, the range that can be determined as the final keyword may be set to be different from each other . For example, if the range that can be selected as the final keyword is set to the first range, the final keyword may be selected based on a relatively strong constraint. If the range that can be selected as the final keyword is set to the second range, the final keyword may be selected based on the relatively weak constraint. Based on these setting values, it is possible to expand and narrow the range of keywords. The setting of the keyword range can be performed according to the user's reading ability (or problem solving ability, learning ability), which can help the user understand fingerprints more stereoscopically.

5 is a conceptual diagram illustrating a key sentence extraction method according to an embodiment of the present invention.

In Fig. 5, a key sensense extraction method of a fingerprint is disclosed.

Referring to FIG. 5, the parser can extract a key sentence in consideration of the understanding of sentences in the fingerprint, the relationship between the sentences in the fingerprint, the sentence weights, and the like.

The parsing device can extract a sentence that is judged as a key sentence in the fingerprint.

The first extraction key sentence 510 may be broadened in the range of extraction, and may be selected according to the priority order, and a plurality of sentences may be provided as the first extraction key sentence 510.

The first extraction key sentence 510 is provided to the key sentence trainers for learning about the key sentence of the fingerprint, and the determination of the key sentence trainer for the first extracted key sentence 510 is performed, The key senses 520 can be determined. Specifically, when the first extraction key sentence 510 has ten sentences, a sentence determined to be actually a key sentence among the first extracted key sentences 510 may be determined as the selected key sentence 520.

The parser may perform the first-order learning based on the result determined by the selection key sentence 520. In the fingerprint for the sentence determined by the selected key sentence 520 among the first extracted key sentences 510 and the sentence determined by the selected key sentence 520 among the first extracted key sentence 510, Understanding the sentence, the relationship between the sentences in the fingerprint, and the differences in sentence weights can be studied.

A determination of the final key sentence 530 for the fingerprint may be performed based on this primary learning result.

Correction information for the final key sentence 530 of the user learning the fingerprint for additional secondary learning may also be added. If it is not the actual key sentence selected from among the provided information selected as the key sentence in the learning result fingerprint of the user, the learning users can input the additional selected key sentence 520, and the additional selected key sentence information The key sentence of the fingerprint can be determined through additional secondary learning based on the fingerprint 520.

The parsing device can extract the key sentence in the fingerprint more accurately based on a plurality of learning results for these keywords.

According to the embodiment of the present invention, the parser can extend the range of the key sentence according to the setting of the user. Key sentences existing in the fingerprint can be prioritized according to importance. As described above, when the final key sentence is determined in consideration of the sentence in the fingerprint, the relationship between the sentences in the fingerprint, the sentence weight, and the number of repetitions of the word, the range that can be determined by the final key sentence is different Can be set. For example, if the range that can be selected by the final key sentence is set to the first range, the final key sentence can be selected based on the relatively strong constraint. If the range that can be selected by the final key sentence is set to the second range, the final key sentence may be selected based on the relatively weak constraint. Based on these settings, it is possible to expand and reduce the range of the key sentence. Such a key sentence range can be set according to the user's reading ability (or problem solving ability, learning ability), and thus it is possible to help the user understand the fingerprint more stereoscopically.

6 is a conceptual diagram illustrating a method for providing keyword information and key sentence information according to an embodiment of the present invention.

In Fig. 6, a method for separately providing keyword information and key sentence information is disclosed.

Referring to FIG. 6, keyword information and key sentence information may be provided in consideration of a speed of a user's reading.

For example, when a fingerprint is composed of a plurality of paragraphs, a plurality of keywords and key senses per paragraph can be extracted according to the user's setting. The user can compare the extracted keywords and key senses after reading the paragraph It is possible to judge the degree of understanding of the fingerprint. The user device can determine the position of the user's eyes and determine the location of the user's current reading fingerprint.

In addition, the user can check the extracted keywords and key sentence after completion of the reading of the entire fingerprint. In addition, in the embodiment of the present invention, information on the relationship between a plurality of sentences constituting the fingerprint can be visually provided. For example, a fingerprint that includes claim 1, ground 2, and ground 5 can be visualized as a different depth / different color of claim 1, ground 2, and ground 5, and provided to the user.

According to the embodiment of the present invention, on the basis of the understanding of the sentence, the requirement to be extracted from the fingerprint based on the sentence including the selected keyword, the relation between the sentence including the selected keyword and the surrounding sentence, Learning can proceed. Thereafter, the relationship between the sentences can be rearranged to the user in three dimensions and visually provided to the user. Based on the information on the relationship between these sentences, the user can improve the reading ability of the fingerprint.

The embodiments of the present invention described above can be implemented in the form of program instructions that can be executed through various computer components and recorded in a computer-readable recording medium. The computer-readable recording medium may include program commands, data files, data structures, and the like, alone or in combination. The program instructions recorded on the computer-readable recording medium may be those specifically designed and configured for the present invention or may be those known and used by those skilled in the computer software arts. Examples of computer-readable media include magnetic media such as hard disks, floppy disks and magnetic tape, optical recording media such as CD-ROM and DVD, magneto-optical media such as floptical disks, medium, and hardware devices specifically configured to store and execute program instructions, such as ROM, RAM, flash memory, and the like. Examples of program instructions include machine language code, such as those generated by a compiler, as well as high-level language code that can be executed by a computer using an interpreter or the like. The hardware device may be modified into one or more software modules for performing the processing according to the present invention, and vice versa.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed embodiments, but, on the contrary, Those skilled in the art will appreciate that various modifications and changes may be made thereto without departing from the scope of the present invention.

Accordingly, the spirit of the present invention should not be construed as being limited to the above-described embodiments, and all ranges that are equivalent to or equivalent to the claims of the present invention as well as the claims .

Claims (11)

Fingerprint and problem description method,
Providing an online education providing device with a fingerprint commentary to a user device; And
Wherein the on-line training provision device provides the problem description to the user device,
Wherein the fingerprint commentary is performed based on a key word and a key sentence of the fingerprint analyzed by the parsing device. The method according to claim 1,
Wherein the keyword is a word central to the reading of the fingerprint,
The key sentence is a sentence centered on the reading of the fingerprint,
Wherein the problem description provides a region of the fingerprint associated with the problem with a separate indication. The method according to claim 2,
Setting a word whose repetition is repeated by a first threshold number or more as a primary keyword;
Setting a word that is repeated by the parser over a second threshold number as a secondary keyword;
Receiving the selected keyword among the primary keyword and the secondary keyword as feedback for the primary keyword and the secondary keyword;
Performing a learning based on a repetition count of the selected keyword, a sentence including the selected keyword, a relation between a sentence including the selected keyword and a surrounding sentence, and a sentence weight; And
And the parsing device is determined based on extracting the final keyword as the keyword on the fingerprint based on the result of the learning. The method of claim 3,
Extracting a first extraction key sentence by the parser;
Receiving the information on the selected key sentence among the first extracted key sentences;
Wherein the parser analyzes the sentence determined by the selected key sentence among the first extracted key sentences and the sentence determined for the sentence not determined by the selected key sentence among the first extracted key sentences Performing a learning on differences in sentence weights, relationships between sentences in the fingerprint, and differences in sentence weights; And
Wherein the parsing device is determined based on the step of determining the key sentence on the fingerprint based on the learning. 5. The method of claim 4,
Wherein the extraction range of the keyword and the key sentence is adjusted according to a setting of a user. An on-line training provision device for performing fingerprints and problem-
A communication unit for communicating with the user device; And
And a processor operatively coupled to the communication unit,
The processor provides a fingerprint commentary to the user device,
Wherein the user device is configured to provide a problem description,
Wherein the fingerprint commentary is performed based on a keyword and a key sentence of the fingerprint analyzed by the parsing device. The method according to claim 6,
Wherein the keyword is a word central to the reading of the fingerprint,
The key sentence is a sentence centered on the reading of the fingerprint,
Wherein the problem description provides an area of the fingerprint associated with the problem along with a separate indication. 8. The method of claim 7,
Wherein the parsing apparatus sets a word repeated at least a first threshold number as a primary keyword, sets a word repeated at least a second threshold number as a secondary keyword, and feeds back the primary keyword and the secondary keyword And a selection unit configured to receive a selection keyword selected from the primary keyword and the secondary keyword and to generate a selection keyword based on a repetition number of the selected keyword, a sentence including the selected keyword, a relationship between a sentence including the selected keyword and a surrounding sentence, And the final keyword is extracted as the keyword on the fingerprint based on the result of the learning. 9. The method of claim 8,
Wherein the syntax analyzing apparatus extracts a first extraction key sentence, feeds back information on a selected key sentence among the first extracted key sentences, and receives a second extracted key sentence from the first extracted key sentence, Learning about the sentence in the fingerprint, the relationship between the sentences in the fingerprint, and the difference in the sentence weight for the sentence not determined by the selected key sentence among the first extracted key sentence And determines the key sentence on the fingerprint based on the learning. 10. The method of claim 9,
Wherein the extraction range of the keyword and the key sentence is adjusted according to the setting of the user. A computer program for executing the method according to any one of claims 1 to 5.

Images