KR102182248B1 - System and method for checking grammar and computer program for the same - Google Patents

Abstract

A grammar checking system includes: a storage module formed to store a word meaning dictionary database including a plurality of words and attribute information corresponding to each of the plurality of words, and processing rule information defined based on the attribute information; an input module formed to extract at least one sentence from input data based on the word meaning dictionary database, and separate at least one word or phrase from the at least one extracted sentence; a grammar checking module formed to detect a grammar error by applying grammar processing rules defined based on the attribute information, in regard to the at least one separated word or phrase; and an output module formed to provide information about the error detected by the grammar checking module. Here, the attribute information is defined step by step by including a first attribute indicating components of words, a second attribute indicating classes of words and a third attribute indicating meaning characteristic categories of words. The grammar checking system can apply accurate grammar rules by using the word meaning dictionary database which reclassifies word classes in detail in accordance with applied grammar rules, and can enable a computer to recognize grammar rules and process grammar checking by expressing each of the grammar rules as a function format by a predetermined protocol.

Description

Grammar test system and method, and computer program for the same TECHNICAL FIELD [SYSTEM AND METHOD FOR CHECKING GRAMMAR AND COMPUTER PROGRAM FOR THE SAME}

The embodiments relate to a grammar checking system and method and a computer program therefor. More specifically, the embodiments may apply accurate grammar rules using a word semantic dictionary database in which parts of speech are reclassified in detail according to the applied grammar rules, and each grammar rule can be applied according to a predetermined protocol. It is about the technology related to the grammar checking system that enables automated processing by expressing it in the form of a function.

Grammar checks, from providing information necessary for the correction of simple spelling and spelling in a narrow way, to check for stylistic errors in order to support effective writing of documents or sentences in a broad sense, provide correct terms/expressions, or punctuation marks. It means to inform the suitability of the use of punctuation marks, etc., or to provide correction information for errors or usage examples of words.

Since most of the document creation work in the modern society is performed by a computer, document creation programs used for document editing, desktop publishing, and electronic forms have also been developed to include a grammar check function. For example, Unexamined Patent Publication No. 10-2008-0032281 discloses a program that automatically converts spaces, spelling, and standard word notation.

However, the grammar check function provided in the conventional document writing program does not sufficiently reflect various uses of Korean grammar, and thus has a clear limitation in that the test result is often inaccurate. In particular, in Korean, grammar rules applied according to the meaning of words of the same part of speech are often different from each other, but grammar errors due to the classification of the meaning of words cannot be detected only by the grammar test through the conventional 5 words and 9 parts of speech classification. Accordingly, most of the retrieval of documents has been done manually by humans.

In recent years, with the development of IT technology, technologies using machine learning or big data are emerging. However, since the machine learning method compares multiple sentences and converts them into similar sentences or words without considering the logical aspect of grammar, it is difficult to guarantee the accuracy of the test. In addition, the technology using big data is limited to constructing standard words and typos one-to-one and changing typos to standard words, so it is difficult to be used for complex grammar tests that require stem analysis.

Unexamined Patent Publication No. 10-2008-0032281

According to an aspect of the present invention, since grammar rules applied according to the meaning of words of the same part of speech are different from each other, the grammar rules applied based on the semantic properties of words beyond the component and part of speech classification of words according to the prior art A grammar check system and method capable of reclassifying identical words in detail and performing a rule-based grammar check based on this, and a computer program therefor may be provided.

The grammar checking system according to an embodiment of the present invention is configured to store a word semantic dictionary database including a plurality of words and attribute information corresponding to each of the plurality of words, and processing rule information defined using the attribute information. Configured storage module; An input module configured to extract one or more sentences from input data, and to separate one or more words or phrases from the extracted one or more sentences based on the word semantic dictionary database; A grammar checking module configured to detect a grammar error by applying a grammar processing rule defined using the attribute information to one or more separated words or phrases; And an output module configured to provide information on an error detected by the grammar checking module.

In this case, the attribute information is defined stepwise including a first attribute indicating a component of a word, a second attribute indicating a part of speech of a word, and a third attribute indicating a semantic property classification of the word.

In one embodiment, the processing rule information includes one or more of a spacing rule, a search use rule, a ending rule, and a word combination rule defined using the third attribute of a word.

In one embodiment, the processing rule information includes one or more grammar processing rules defined in the form of a processing function executed on condition of one or more of a document type, a sentence position, and an adjacent character.

The grammar checking system according to an embodiment is configured to generate a compound verb in which the conjugation is defined by combining a basic verb with a conjugated type defined in advance with other words to expand the stem, and to store the compound verb in the word semantic dictionary database. It further includes a term management module.

A grammar checking system according to an embodiment includes: a typo generator configured to generate a predicted typo by substituting a part of a word included in the word semantic dictionary database by using a preset group of letters or words; And a typo checking module including a typo checking unit configured to perform a typo check for the separated one or more words or words based on the predicted typo generated by the typo generating unit.

A grammar checking method according to an embodiment of the present invention includes storing a word semantic dictionary database including a plurality of words and attribute information corresponding to each of the plurality of words in a grammar checking system; Storing processing rule information defined by using the attribute information in the grammar checking system; Extracting, by the grammar checking system, one or more sentences from input data; Separating, by the grammar checking system, one or more words or phrases from the extracted one or more sentences based on the word meaning dictionary database; Detecting, by the grammar checking system, a grammar error by applying a grammar processing rule defined using the attribute information to one or more separated words or phrases; And providing, by the grammar checking system, information on the detected error.

A grammar checking method according to an embodiment comprises the steps of: generating, by the grammar checking system, a compound verb having a conjugated form defined by combining a basic term with a pre-defined conjugated form with another word to expand a stem; And storing, by the grammar checking system, the compound verb in the word meaning dictionary database.

A grammar checking method according to an embodiment includes the steps of: generating, by the grammar checking system, a predicted typo by substituting a part of words included in the word semantic dictionary database using a preset letter or group of words; And performing, by the grammar checking system, a typo check for the separated one or more words or phrases based on the generated predicted typo.

In one aspect of the present invention, a computer program is combined with hardware to execute the grammar checking method according to the above-described embodiments and may be stored in a computer-readable recording medium.

A grammar check system and method according to an aspect of the present invention builds a word semantic dictionary database in which parts of speech are reclassified in detail according to applied grammar rules, and grammar rules for each word and detailed type of part of speech in the word semantic dictionary database By registering, there is an advantage in that even words of the same part of speech can accurately check grammar by different grammar rules applied according to their meaning.

In addition, in the grammar checking system and method according to an aspect of the present invention, since the grammar rules are expressed in the form of a function according to a predetermined protocol, the grammar knowledge of an expert is transmitted so that the computing device can understand and process it, and through this It is possible to perform a grammar check in an automated way, and there is an advantage that grammar rules can be easily modified and added.

In addition, in the grammar checking system and method according to an aspect of the present invention, by automatically generating an extended stem of a single word and registering the extended stem for a combined word and/or a compound word in the word semantic dictionary database based on this, There is an advantage in that the database is automatically updated to include the compound verbs to be used, allowing accurate grammar checks.

Furthermore, in the grammar checking system and method according to an aspect of the present invention, by creating a list of typos for each letter and word in which typos may occur, and automatically generating and storing typos for each standard word using this, If there are typos, there is an advantage that it can provide a function to convert the typos into standard words.

1 is a block diagram of a grammar checking system according to an embodiment.
2 is a flow chart showing each step of a grammar checking method according to an embodiment.
3 is a conceptual diagram illustrating an example of detailed classification of parts of speech by a grammar inspection method according to an exemplary embodiment.
4A and 4B are conceptual diagrams of an exemplary user interface showing grammar items according to detailed classification of parts of speech in the grammar checking system according to an embodiment.
5A to 5C are conceptual diagrams of an exemplary user interface illustrating a grammar processing rule by a function in a grammar checking system according to an exemplary embodiment.
6 is a flowchart illustrating a process of generating a compound verb through stem expansion in a grammar checking method according to an exemplary embodiment.
7 is a conceptual diagram illustrating a process of generating the compound verb shown in FIG. 6.
8A and 8B are conceptual diagrams of an exemplary user interface related to generation of compound verbs through stem expansion in a grammar checking system according to an embodiment.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the drawings.

1 is a block diagram of a grammar checking system according to an embodiment.

Referring to FIG. 1, the grammar check system 3 according to the present embodiment includes a storage module 10, an input module 20, a grammar check module 30, and an output module 60. In one embodiment, the grammar checking system 3 further comprises a term management module 40. Also in one embodiment, the grammar checking system 3 further comprises a typo checking module 50. In addition, each of the modules 10, 20, 30, 40, 50, 60 described above may include one or a plurality of functional units.

The devices described herein may be entirely hardware, or may have aspects that are partly hardware and partly software. For example, the grammar checking system 3 and each system, device, server, and each module or unit included therein communicate with the grammar checking system 3, a device for sending and receiving data of a specific format and content in an electronic communication method. And software related thereto. In this specification, terms such as “unit”, “module”, “server”, “system”, “platform”, “device” or “terminal” are intended to refer to a combination of hardware and software driven by the hardware. do. For example, the hardware here may be a data processing device including a CPU or other processor. In addition, software driven by hardware may refer to an executing process, an object, an executable file, a thread of execution, a program, and the like.

In addition, in the present specification, each unit constituting the grammar checking system 3 is not intended to refer to a separate component that is physically separated. That is, in FIG. 1, each of the modules 10, 20, 30, 40, 50, 60 of the grammar checking system 3 is shown as separate blocks that are distinguished from each other, but this allows the grammar checking system 3 to be executed by It is functionally classified by operation. Depending on the embodiment, some or all of the above-described modules may be integrated into the same single device, or one or more modules may be implemented as separate devices that are physically separated from other modules. For example, each of the modules 10, 20, 30, 40, 50, and 60 of the grammar checking system 3 may be components that are communicatively connected to each other in a distributed computing environment.

The grammar checking system 3 is configured to perform a grammar check on input data received from the user device 1 and provide the result to the user device 1. In addition, in one embodiment, the grammar checking system 3 may collect and store information related to a word database, proper nouns, new words, etc. from the external server 2. For the above operation, the grammar checking system 3 may communicate with the user device 1 and/or the external server 2 through a wired and/or wireless network. In this specification, the communication method through a wired and/or wireless network may include all communication methods capable of networking an object and an object, and is not limited to wired communication, wireless communication, 3G, 4G, or other methods. .

In Fig. 1, a user device 1 providing input data for a grammar check is shown in the form of a notebook computer. However, this is exemplary, and as the user device 1, any computing device such as a mobile computing device such as a smartphone, a personal computer, a notebook computer, a tablet computer, and a personal computer (PC) may be used. Can be, and is not limited to a specific type of device.

In addition, in the embodiments described herein, the grammar checking system 3 is implemented as a separate device (eg, a server) that is separate from the user device 1, and receives and inputs input data from the user device 1 The result of the grammar check on the data is provided to the user device 1. However, this is exemplary, and in other embodiments, the grammar checking system 3 itself may be implemented in the form of a user device such as a smartphone, a notebook computer, or a PC, and in this case, a separate user device 1 may be omitted. .

The storage module 10 is configured to store a word meaning dictionary database (DB). The word semantic dictionary DB of the embodiments includes a plurality of words and attribute information corresponding to each word. In addition, the storage module 10 may store processing rule information defined by at least partially using attribute information of each word, and the processing rule information may include one or more grammar processing rules.

In one embodiment, the attribute information of each word is a first attribute indicating a component of a word (e.g., body language, verb, modifier, relational language, affix, independent language, etc.), part of speech (e.g., pronoun, noun, rhetoric, idiom, It includes a second attribute indicating adverb, survey, ending, prefix, suffix, interjection, etc.), and a third attribute indicating semantic classification of words. In addition, the first to third attributes described above may be defined in stages. At this time, the third attribute representing semantic property classification is that grammar rules applied to words among a plurality of words are assigned so that words that are identical to each other are grouped into the same group.

The input module 20 receives input data from the user device 1, extracts one or more sentences from the input data, and separates one or more words or phrases to be subjected to a grammar test again from the extracted sentences. do. In one embodiment, the input module 20 includes a sentence extracting unit 21 for extracting a sentence from input data, and a word extracting unit 22 for separating a word or word from the sentence extracted by the sentence extracting unit 21 Includes. In this case, the process of separating a word or word may be performed based on words included in the word semantic dictionary DB stored in the storage module 10.

In this embodiment, the input module 20 is described as receiving input data from the user device 1, but when the grammar checking system 3 itself is implemented in the form of a user device, the input module 20 It may be to perform a function of receiving input data of a user through an input means (not shown) such as a keyboard and a mouse provided in the system 3.

The grammar check module 30 serves to detect a grammar error by applying a grammar processing rule defined in the storage module 10 to one or more words or words separated by the input module 20. In this case, the grammar processing rule may include grammar items defined for each word and/or part of speech, and may also include a function defined by a predetermined protocol.

In one embodiment, the grammar checking module 30 may include a function management unit 31 for managing and executing functions defined in a grammar processing rule. In addition, the grammar processing rule may include grammar items defined for each word (or phrase) or detailed classification of parts of speech in addition to the function type, and the grammar check module 30 may include a grammar rule application unit for each word ( 32) and a part of speech-specific grammar rule application unit 33 may be included. A process of applying the inspection and grammar processing rules by the grammar inspection module 30 will be described in detail later.

The output module 60 serves to provide information on an error detected by the grammar checking module 30 to a user. In the embodiment of FIG. 1, the output module 60 may transmit information on the error to the user device 1 so that the information can be displayed to the user through the user device 1, but this is only an example, and another embodiment In the case where the grammar check system 3 itself is implemented in the form of a user device, the output module 60 displays information on grammar errors through an output means (not shown) such as a display device provided in the grammar check system 3 You may.

The term management module 40 creates a compound verb with a conjugated form defined by combining a basic term with a conjugated form defined in advance among the words stored in the word semantic dictionary DB of the storage module 10 with other words to expand the stem, This can be added to the word semantic dictionary DB. By automatically defining the basic and conjugated forms of compound verbs based on the basic verbs and adding them to the word semantic dictionary DB, a more accurate test can be made.

In addition to the grammar rules defined in the grammar checking module 30, the typo checking module 50 may perform a typo check for words or words separated by the input module 20, and the typographical check module 50 The typo check result may be provided to the user by the output module 60 as well as information on grammar errors.

In one embodiment, the typo checking module 50 not only determines the presence of typos and/or correction content based on the typos for each word defined in advance, but also through a combination of words or letters for which typos are expected to occur. You can also automatically generate typos for each word. For the above operation, the misspelling detection module 50 automatically generates a misspelling detection rule through a combination of letters and/or words and a misspelling detection unit 51 that detects errors based on the misspellings for each word defined in advance. It may include a typo generator 52 added to the word semantic dictionary DB.

2 is a flow chart showing each step of a grammar checking method according to an embodiment.

Referring to FIG. 2, first, a user may input text to be subjected to a grammar check (S1). Text can be received by the grammar checking system by being inputted to the user device and then transmitted to the server-type grammar checking system, or if the grammar checking system itself is implemented in the form of a user device, the text may be directly input to the grammar checking system. have.

The grammar check system may extract one or more sentences from the input text (S2). The extraction of sentences may be performed by detecting specific characters such as periods (.), question marks (?), and exclamation marks (!) from the content input by the user, or specific typesetting codes such as the enter key. In addition, by detecting the part that ends with the ending (e.g., hearing type, exclamation type, imperative type, narrative type), the sentence is extracted (in this case, the tubular sentence type, noun type, and concatenation type ending are excluded), or/or The beginning and/or end of a sentence by detecting Roman characters, section/chapter/verse/gwan, article/paragraph/number/tree/subject/detail, or special characters (eg, □, ☞, ｏ, -, etc.) You can also extract a sentence by specifying.

Next, the grammar checking system may separate a word or a word from the extracted sentence based on the words registered in the word meaning dictionary DB (S3). In addition, the grammar test system may compare the separated words or phrases and words registered in the word semantic dictionary DB to determine the detailed classification (ie, semantic property classification) of the words included in the separated words (S4). ). That is, the grammar checking system may search for a separated word or a word included in the separated word from the word semantic dictionary DB, and specify a detailed classification corresponding thereto.

3 is a conceptual diagram illustrating an example of detailed classification of parts of speech by a grammar inspection method according to an exemplary embodiment.

Referring to FIG. 3, in the grammar checking method according to the embodiments, each word may have attribute information defined by arranging a first attribute 31, a second attribute 32, and a third attribute 33 in stages. have. The first attribute 31 represents a component of a word (eg, body language, average, modifier language, relational language, affix, independent language, etc.), and in the example of FIG. 3, an example in which the first attribute 31 is a body language is illustrated. In addition, the second attribute 32 represents the part of speech of a word (eg, pronoun, noun, rhetoric, tubular sentence, adverb, investigation, ending, prefix, suffix, interjection, etc.).

At this time, the grammar checking method according to the embodiments is differentiated from the conventional spell checking programs in that a third attribute 33 representing semantic characteristics is assigned to a word. The third attribute 33 is a group of words having the same applied grammar among words corresponding to the same component and part of speech, by parts of speech. For example, pronouns among verbs can be classified into semantic properties such as space, objects, people, work, and time according to the grammar applied thereto, and rhetoric among pronouns can be divided into semantic properties such as number order and quantity. . By applying the third attribute 33 defined as described above, attribute information of a word representing time among pronouns that is both a body language and a pronoun may be defined in a form that can be read by a computing device, such as [*cement-pronoun-time*]. In this case, [* and *] are symbols indicating the start and end of attribute information, but the shape of the symbol is not limited thereto.

In an embodiment, the third attribute 33 may be further divided into one or more upper attributes 331 and 332 and one or more lower attributes 333 representing detailed classification of each of the upper attributes 331 and 332. For example, the upper attribute 332 called time may have one or more lower attributes 333 such as time date, period, time time, year month and month, and the third attribute 33 of each word is such a lower attribute 333 ) May be further used. For example, attribute information of a word that is a noun among body speech and has semantic properties corresponding to the time point may be defined as [*At the time of body speech-noun-time-time*].

The third attribute 33 shown in FIG. 3, that is, semantic classification, is exemplary, and does not cover all types of third attributes defined for each word in the grammar checking system and method according to the embodiments. This will be readily understood by those skilled in the art. That is, in the embodiments of the present invention, the third attribute (i.e., detailed classification of parts of speech or semantic property classification) may be defined in an arbitrary manner as long as the grammar to be applied to words is identical to each other, or words corresponding to each other can be grouped. And, it is not limited to the example shown in FIG. 3.

Referring back to FIG. 2, in an embodiment, a detailed classification of the part of speech of a word may be determined by collecting homophones or words among words registered in the word semantic dictionary DB and comparing them with the target word (S5). This is, for example, in the case of words that have the same or similar pronunciation, such as "Payment" and "Payment", or that are prone to typos, even if the input text contains "Payment", the word "Payment" is a homophone. It is to accurately determine the semantic property classification of the word by searching together in the semantic dictionary DB.

When the detailed classification of the components, parts of speech, and parts of speech of a word is determined, the grammar check system applies the grammar processing rules predefined in the grammar check system according to the attribute information of the specified word, and thus the word or word containing the word is It is possible to detect grammatical errors related to (S61, S62, S63 and S7). The process of detecting grammatical errors can be divided into a plurality of steps according to the nature of the applied processing rule. This is, for example, words that designate typos, spaces, and pastes for each word included in each word semantic dictionary DB. One of a step (S61) of applying a grammar item for each word, a step of applying a grammar processing rule described in the similar word function for each word (S62), and a step of applying the grammar processing rule described in the sentence error function for each word (S63) It may include more than one.

In embodiments of the present invention, the grammar processing rule to be applied in the step (S61-63) of applying each of the grammar processing rules described above may be at least partially using the detailed classification of the part of speech of the word (that is, attribute information of the word). have. That is, according to the detailed classification of the word's part of speech, grammar items for each word, such as typos, spaces, and pasting, are defined, or the execution condition or execution content of a function for applying grammar processing rules for the word is detailed Can be defined based on classification. In addition, the process of applying the grammar processing rule based on the detailed classification of parts of speech may be performed independently from the step S7 of applying the grammar processing rule for each word described above.

The process of applying the grammar processing rule in the grammar checking system and method according to the embodiments may be performed at least partially using a processing function. In the present specification, the processing function is a description of the contents for specifying an error word or word and/or changing it to a correct word or word in the form of a character string readable by a computing device based on a predetermined protocol. Refers to.

Application of a grammar rule by a grammar checking system and method according to embodiments will be described in detail below with reference to FIGS. 4 to 8.

4A and 4B are exemplary user interfaces for display to an administrator of a grammar checking system according to an embodiment, and are screens showing grammar items according to detailed classification of parts of speech.

First, referring to FIG. 4A, on the manager screen of the grammar inspection system, an area 1001 in which detailed types of parts of speech related to a target word (ie, attribute information) are displayed, and an area 1002 in which grammar items for the corresponding word are displayed. , An area 1003 in which grammar processing rules are displayed in a function form and an area 1004 in which typos processing rules are displayed may be displayed. In the example of the drawing, the word "value" has attribute information of [*noun-indicated shape-value cost rate*], and the types and divisions of these attributes may be displayed in the area 1001 on the screen.

In addition, grammar matters related to words displayed in the area 1002 correspond to words having attributes defined in the area 1001, such as spacing rules, survey usage rules, ending rules and/or word combination rules, etc. Defines For example, in the area 1002, two check boxes in the [Different Part of Speech] section indicate whether the left and right sides of the target word should be pasted with the words of different parts of speech, respectively, and in front of [Investigation] in the [Term] section. Two checkboxes indicate whether the left and right of the target word should be examined and pasted respectively. Since these spacing rules or survey usage rules may vary depending on the semantic classification of the word (ie, "value cost charge" in this example), in the embodiments, the semantic classification of the word is designated as an attribute and It defines the grammar rules.

Further, the area 1003 represents, in the form of a function, a processing rule executed under conditions such as the type of document on which the grammar check is performed, the position of a sentence including the target word, and/or adjacent characters of the target word. In the area 1003 of FIG. 4A, the (processing) function corresponds to a grammar processing rule according to the similar word function for each word performed in step S62 of FIG. 2, and the (sentence) function is performed in step S63 of FIG. It corresponds to the grammar processing rule according to the sentence error function for each word.

In an embodiment, the grammar processing rules may define processing conditions or contents by using at least partially attribute information related to a word. For example, the “■[*cheese*] value ▶[*cheese*] value” shown in the area 1003 is, if the attribute information of the word preceding the target word “value” is [*cheese*] If the word and "value" are spaced, it means that they are corrected to be pasted. Details of the function will be described later with reference to FIGS. 5A to 5C.

On the other hand, the area 1004 is a part for instructing a typo detection rule by using an error notation defined in advance for a corresponding word, and is processed by the typo checking unit 51 of the typo checking module 50 shown in FIG. Indicates the changes that are made. That is, when a word indicated as a typo in the area 1004 is extracted from a sentence, the grammar checking system may check the typo detection rule as shown in FIG. 4A and convert the typo into a standard word. In addition, in applying a grammar processing rule for a word, a grammar check may be performed by first converting the word determined as a typo into a standard word, and then applying a grammar processing rule for the standard word.

On the other hand, the typo generator 52 of the typo checking module 50 shown in FIG. 1 automatically generates a typographical character to be indicated through the region 1004 of FIG. 4A using similar letters or words, and checks the grammar. Can be stored in the system. To this end, the typo generator 52 generates a group of letters or words in which typos can easily occur between each other, and if any one letter or word of the typo group is included in the test target word, Alternatively, you can automatically generate expected typos by replacing them with words.

For example, groups of words that are prone to typos are as follows: ``Gun/Chun, Gender/Hyung, Gil/Jil, Oil/Diameter, Jang-Yi/Yun, Dong-Yi/Dong-Y/Etc, Tung-Yi /Tongi, Sungi/Songi, Kick/Reap, sweet/Look. Hairy/Treat/Tri, seemingly/desire, in/out, fun/jami, roll/reap, butch/paste'. In addition, groups of letters that are prone to typos among each other include the following: 『dog/he/ge/gye, loss/?/pe/lung, rae/?/re/rye, mae/me , New/Three, My/Yes, Sre/Sle, Comb/Debt/Light, F/F, F/F/F, F/F, F/F, F/F, Show/Small, On/Foo, C/C , E/up, ki/ear, tic-tick/tu-tu, ear/female, shea/she, subject/give, boa/see, couple/do, do/do, oh/oh, shit/wa, a/ㄲ/ㅋ , ㄷ/ㄸ/ㅌ, ㅅ/ㅆ, ㅈ/ㅉ, ㅂ/ㅃㅍ/ㅎ, E/Ye/Ae, Lee/Up, Why/Etc, Room/Pak, Gol/Gore (from Jongseong to Lee Eul (ㄹ) Eligible example), Bow/hal, Kkak/cut, get/go, number/number, up/laugh/up』.

As an example, explaining the process of automatically generating typos for chocolate, the grammar check system in this embodiment includes cho/cho/cho, co/cur/curl/call, rat/?/let/rock/? Based on the typo group of /?/?/?/lit/?/rat/?t/let/ryet/lit, you can create a predicted typo of "chocolate". That is, since cho/cho/cho belong to the same typo group, the expected typos “chocolate” and “chocolate” are expected by replacing “cho” with “cho” and “cho” belonging to the same typo group in “chocolate”. Can be created. The same principle can be applied to other letters of "chocolate", as a result, the letter that can replace "second" is "second" and two other letters in the same group, and the letter that can replace "call" is "coll." Since there are 3 other characters in the same group as "Lit" and 14 other characters in the same group as "Lit", a total of 70 (2×3×14) expected typos are automatically generated. And it is possible to store the conversion rule for this in the grammar check system.

In addition, in an embodiment, in addition to a method of replacing letters using a group of typos, typographical characters may be generated based on grammar rules applied based on adjacent letters. As an example, a typo can be created by adding a gap between the front word and the back word for a word whose front word ends as a vowel while two words are combined among words registered in the word semantic dictionary DB. Same as: 『Gulbi-Jjigae (Othalja: Oyster Bit Jjigae), Kimchi-Jjigae (Othalja: Kimchi-jjigae), Kkakdugi-Jjigae (Kkakdugijigae), and Saury-Kimchi-Jjigae (Otalja: Saongchitkimchijigae). Or/or, in the case of a word that contains a gap between the preceding word and the back word while two words are combined, a typo may be created by removing the gap sound, and examples of this are as follows: day)".

4B is another exemplary screen showing grammar items according to detailed classification of parts of speech.

Referring to FIG. 4B, on the manager screen of the grammar checking system, an area 1011 in which detailed types of parts of speech related to a word (ie, attribute information) are displayed, and an area in which ending rules are displayed as grammar items for the word ( 1012), an area in which spacing and investigation usage rules are displayed (1013), an area in which grammar processing rules are displayed in a function form (1014), an area in which a word combining rule is displayed (1016), and an area in which a typo handling rule is displayed (1015) ), etc. may be displayed. In order to avoid duplication of description, only a portion of each region shown in FIG. 4B that is different from the user interface described above with reference to FIG. 4A will be described.

4B shows a grammar processing rule defined for the word "because" in which part of speech corresponds to a ending and semantic property classification as a term is classified as a connection type according to the present embodiment. At this time, the type of proverb to which the ending “because” can be combined as a grammar processing rule is defined through the area 1012, and in the example shown in FIG. 4B, the word is a rule using the ending ending “because” It indicates that the grammar processing rule is defined so that it can only be combined with the basic form rule stem of a verb that follows the transformation, and cannot be attached to the basic form irregular stem of the verb. For example, if "because" is combined with "make", which has different basic types (rule: make, irregular: make, honorable name: make) in which the shape of the rule deformation and the irregular deformation are different, then "because of making" is correct, "Because we make" is an incorrect word, and this combination rule is defined by [Types of endings] in Fig. 4B.

In addition, the region 1016 of FIG. 4B is a combination rule of the word “because”, and when “because” is combined with a word ending with a combination of vowels and rieul, grammar processing rules are defined to be combined in the form of a substitute word “because” Indicates that

Meanwhile, as shown in FIGS. 4A and 4B, the grammar checking system according to the embodiments is configured to execute a grammar processing rule by executing a (processing) function or a (sentence) function defined for a word. At this time, each function is defined by a predetermined protocol to convert grammatical expertise into a language that can be read by a computing device, and converting a word or letter using the detailed types of parts of speech defined in the word semantic dictionary DB For.

Each symbol used in a protocol for defining a function in the grammar checking system according to an embodiment, and a description of the name and function thereof, are given in Tables 1 to 3 below.

sign name Description and use case ■ Change Refers to the term to be reviewed for grammar rules ▶ Recommendation Means the correct term/expression to change the target term exit Exit While processing specific grammar rules, forcibly exiting the processing routine
Command to come out
Ex) exit: exit unconditionally,
▶exit Exit without change [*Part of speech location*] Type of speech The location of the group where parts of speech are divided in detail, that is, the detailed types of
Means. Example) [*corpse-noun-dependence-time*], [*verb-adjective
-Connected type-honor*], [*adjective-like*] ※ Comments (description) Help on grammar rules or how to write grammar rules If Sentence condition If a sentence contains a specific word, process it In if condition scope Symbol used to set the conditional range of If statement =, !=
and, or, xor Equal sign/logical
Condition =(positive condition), !=(negative condition)
and(intersection), or(union), xor=(exclusion condition) docu Document type This is a variable that identifies and designates the type of document.
Type of document = {laws, presidential decrees, enforcement regulations, precedents,
Interpretation of laws and regulations, research papers, thesis, officials, plans}
Example) ■[*word*] If docu ='President'
▶[*words*]
※ If this document is'President',
Command to modify the expression to'I like it' Contents Document location Designate the current location among the documents to be reviewed
Contents = start (start), end (end), page (page),
para (paragraph), sen (sentence), word (word)
before/after (words and sentences immediately before/after )
pre/post (from start/end to current position)
Example) ■AAA if local government in pre sen ▶BBB
※ From the beginning of the sentence, before the current word
When the term'local government' is included,
Command to change'AAA'to'BBB'

sign name Description and use case [] One-to-one matching
option The order of the words to be replaced and the suggested words matched
Option to match words/phrases/phrases one-to-one (1:1) <> Choice option Survey selection option to choose according to the end of the word syllable {} Many-to-one matching
option Option to group multiple words together | Word pipe One-to-one optional item separator to separate words/phrases ; Command execution Means to change the target word and execute the following command ^ Blank Means blank, blank is possible in sentences & text 1 character (show character) ~ String One or more words refer to a letter/word, and words to change and recommendations
Must match each other in order ~~ String Must mean more than one string - 1 or more
String or
If not No or more than one string @ Special Characters Special character symbol (show special character) $ number A number consisting of 1 number (show the number) # Numeric string It refers to a sequence of more than one number, and the words to be replaced and recommended words are in order.
Must match each other accordingly - Negative words Among the results detected before the sign "-",
The term is not converted.
Ex) Search word-[words excluded],
However, negative words must contain the words to change / OR item
Separation option OR selection option item separator to separate sentences

sign name Description and use case → Description or
address Description of the conversion rule provided to the user, eg) →"Once" /n Enter Meaning the end of a sentence /nl End of paragraph Meaning the end of the paragraph ¶ Forced line sharing New line Int ■ int(#, 3) ▶ $$$,$$$,$$$ Decompose natural numbers by 3 digits
int(#, 0) do not decompose natural numbers Float ■ float(#, 3, 2) ▶ $,$$$,$$$.$$ 3 natural numbers
Truncate to 2 decimal places
Ex) float(#, 3, 2) 34,631,965.73 Built-in function
(letter) Phonetic function First letter last (letter), Last first letter (letter), First letter (letter),
Back character (character) Built-in function
(word) Word function Front word (word), trailing word (word), preceding phrase (word),
Back sentence (word), front sentence (word), back sentence (word) collection Word end meaning A word whose finality ends with a vowel.
E.g.) ■If the car is the initial character, it is a vowel ▶ car Consonant Word end meaning Words that end with a consonant Yul/Rhythm Word end meaning Vowels, words ending in b,
Vowels, words ending with letters other than ㄴ Voiced/unvoiced Word end meaning Voiced sound (vowel, ㄴ, ㄹ, ㅁ, ㅇ),
Unvoiced (vowels, letters other than ㄴ, ㄹ, ㅁ, ㅇ) Korean, English Word end meaning Words ending in Korean, words ending in English Chinese/Simplified Word end meaning Words ending in Chinese characters, words ending in Simplified Chinese Front and back sentences Type of sentence Statement, question, exclamation, imperative
(Yes, if before sen = question) Word ingredient Sentence component Subject, object, indirect object, predicate, adverb, bore
(Yes, if before sen = question)

5A to 5C are conceptual diagrams of an exemplary user interface illustrating a grammar processing rule by a function in a grammar checking system according to an exemplary embodiment.

First, FIG. 5A is a screen showing details of the function “because/because” defined for the word “because” shown in FIG. 4B. As shown in Fig. 5A, the function of "because / as" is detected as its execution condition, "as in [*terms-noun form*]" or "as in [*noun*]" (condition 1), or It is executed in response to the detection of a word of 『[*Terminology-Basic Type*]Because』" (Condition 2), and changes to be executed at this time are also defined based on the attribute information of the word. For example, the function written in the first line of condition 1 changes the grammar processing rule to change it to “because it is [*noun-person relationship*]” when a word consisting of “[*noun-person relationship*]” is detected. Specify.

As shown in Fig. 5A, even if the function is executed under the same condition, attribute information of the target word or the word included in the target word (for example, [*noun-person relationship*], [*noun-organization and building*] , [*Noun-indicative shape*], and [*noun-abstract concept*], detailed grammar processing contents may be different, so in embodiments, grammar processing rules based on attribute information of words are It is defined as a form and can be used to automatically perform text inspection.

On the other hand, the function "because / as" of FIG. 5A corresponds to correction of typos in "because" and "as" among grammar processing rules, and FIG. 5B is another example of "cannot/cannot" related to spacing in grammar processing rules. Represents a function of ".

Referring to FIG. 5B, in the function of “cannot/cannot”, a word of “not [*do*]” or “wrong [*do*]” as its execution condition is detected (condition 1), or Executed in response to the detection of a word of [*haha*]』 or 『wrong[*haha*]』 (condition 2), and the change to be executed at this time is defined based on the attribute information of the word. For example, the function written in the first line of condition 1 is that a sentence containing "is" or "to" is combined with "[*terms-basic form*]ji", followed by "not [*do*]". In this case, the grammar processing rule that removes the spacing error in the 『Can't [*hada*]] part and changes it to 『Non[*hada*]』 is specified.

Meanwhile, in another embodiment, as illustrated in FIG. 5C, the function itself targets specific attribute information, and the function may be executed on condition that a phrase including a word having the attribute information is detected. The function "cheong-noun-unit-time light speed" function shown in FIG. 5C is "liver", "currency", "root", after a word including [*unit-time light speed*] as attribute information. It is executed in response to a search for a word combining words such as "large", "distance", "only", "ban", "bird", "car", "chi", "female", and a [ A grammar processing rule that inserts or removes a space between a word having the property of *unit-time speed*] and the letter following it is specified (conditions 1 to 3).

In addition, a function representing a grammar processing rule may be defined based on attribute information of words adjacent to each other, and condition 4 of "Chong-on-noun-unit-time light speed" shown in FIG. 5C is [*unit-time light speed When a word with the attribute of *] and a word with the attribute of [*dependence-position direction*] are combined, a grammar processing rule is specified to insert a space between the combined words.

However, the form of the function of the grammar processing rule described above with reference to the drawings is exemplary, and the types of grammar processing rules and functions that can be executed by the grammar checking system and method according to the embodiments are as examples described in the present specification. It is not limited. That is, as shown in Figs. 4A and 4B, a spacing rule, a search use rule, a ending rule and a word combination rule, etc., defined for a word having a specific attribute, or a specific attribute as shown in Figs. 5A to 5C. In relation to a word, processing rules defined based on conditions such as the type of document including the word, the position of the sentence, and adjacent characters of the word may all correspond to the grammar processing rules of the present specification.

Referring back to FIG. 2, in the grammar checking method according to the embodiments, by checking whether an error is detected in a specific word in the input text or a word containing the word through the above process (S8), when an error exists. Information on the grammar error may be output as a result of the inspection (S10). In this case, if a substitute word for correcting a corresponding error is defined in the grammar processing rule, the grammar checking system may convert the error part according to the grammar processing rule before outputting the check result (S9). A grammar check can be performed for each sentence in the input text in the manner described above.

In an embodiment, the grammar checking system may automatically generate compound verbs through expansion of stems apart from the above-described grammar check process, and add the resulting compound verbs to the word semantic dictionary DB.

This is to solve the problem that it is difficult to apply grammar processing rules for verbs, adjectives, etc., since the form used in the actual sentence is not limited to the basic form even if the term is registered in the word meaning dictionary DB. For example, even if the ending is the same, the stem shape may have different grammatical characteristics. The ending "yo" cannot be combined after the basic form "do", and only the endings of the concatenated forms "been" and "become". There are cases in which the rules of grammar processing differ depending on the shape of the stem, such as that there is. In the present embodiment, a more accurate grammar check can be performed by automatically generating a compound verb in a form in which the stem of the verb is expanded in a form that is actually used in a sentence.

6 is a flowchart illustrating a process of generating a compound verb through stem expansion in a grammar checking method according to an exemplary embodiment. The operation shown in FIG. 6 may be performed by the term management module 40 described above with reference to FIG. 1. have.

Referring to FIGS. 1 and 6, first, the basic verb management unit 41 of the term management module 40 may collect a single word term from words stored in the word semantic dictionary DB of the storage module 10 (S11). A single word means a word consisting of only one morpheme when the ending is excluded from the morpheme, and, for example, "make" and "hada" correspond to a single word.

Next, the basic verb management unit 41 extracts the same portion as the basic verb in the single word regardless of the change in the ending (S12). For example, in the case of "make" and "hada", the "create" and "lower" parts corresponding to the stem may correspond to basic verbs.

Next, the basic verbal management unit 41 may determine an irregular type of the basic verb and expand the stem of the basic verb (S13). This may be accomplished by using irregular types designated by the administrator of the grammar test system, or may be performed in a manner in which the administrator inputs some or all of the extended forms of the basic verbs and stores them in the text test system.

Next, the compound verb generation unit 42 of the term management module 40 can generate a compound verb corresponding to a combined word or a compound word by expanding the stem by combining the extended stem of the basic verb with other verbs. Yes (S14). In other words, by combining other words in front of the stem of a single word, it is possible to create a compound verb corresponding to a compound word or a combined word. At this time, by generating a compound word or a combined word for the expanded stem of a single word, it is possible to generate a compound word or a combined word in which the stem is expanded. The complex terminology generated as described above may be registered in the word meaning dictionary DB of the storage module 10 (S15).

7 is a conceptual diagram illustrating a process of generating the compound verb shown in FIG. 6.

Referring to FIG. 7, in the case of a single term “hada”, “female” is a word having irregular characteristics, and a basic type (rule: ha, irregular: ha, honorable name: ha), progressive type (ha, ha), present type (han, Haha), connected type (Ha, Ha, Ha Ha, Ha Ha), past type (Ha, Ha, Ha, Ha Ha), future type (Ha, Ha Ha), ending type (Ha, Ha Ha), noun type (Ha, Ha Ha), etc. Is expanded. The grammar checking system according to the embodiments may receive information on irregular characteristics thereof and information on each shape of an extended stem from an administrator using "hada" as a basic term.

Next, the text inspection system according to the embodiments may generate an extended stem of the compound word “like”, which is a compound verb, by combining the word “like” in front of the extended stem of the basic verb “hada”. In other words, the combined terminology of the extended stems of "hada" with irregular "female" in "like" is a basic form (rule: like + ha, irregular: like + ha, honorable title: like + ha), progressive (like) +Do, Like+Have), Present Tense (Like+Han, Like+Have), Connected Type (Like+Do, Like+Do, Like+Do, Like+Do), Past Tense (Like+Ha, Like+Ha, Favor+) The stem can be extended to the noun type (like + like, like + like), future (like + like, like + like), ending (like + like, like + like), noun (like + like, like + like).

On the other hand, in the process of generating compound verbs by the above process, the basic verbs are the basic form (rule: wild, irregular: de, honorable name: eat), continuous type (rule: lift, honorable name: eat), present form (rule: all) , Honorable name: eat), connection type (rule: listen, honorable name: eat), and the like, may be words that have different forms of rule transformation and irregular transformation, and based on this, "make-to", "reduce-to", "power- Complex verbs such as "to hold", "sleep-to-hold", "bounce-to", "to shake-to", "to-to", "to fold-to", and "to-to" may be generated. For example, the compound verb "make-make" is the basic form (rule: make, irregular: make, honorable name: make), continuous (rule: make, honorable name: make), present tense (rule: made, honorable: made), And a connection type (rule: make, honorable name: make).

8A and 8B are conceptual diagrams of an exemplary user interface related to generation of compound verbs through stem expansion in a grammar checking system according to an embodiment.

As shown in Fig. 8A, the basic form (no), irregular (no), progressive (no), present (no), connected (no), past (no, no), future type for the basic term "no" (None), Termination (No), Noun (None), and the forms of honors corresponding to each of these (No, No, No, No, No, No, No, No, No, No) It can be determined by extending the stem of the basic verb. For example, the manager of the grammar checking system may input each form of extended stem for basic verbs among the words recorded in the word meaning dictionary DB.

At this time, the grammar checking system can automatically generate a compound verb "nothing" by combining "A" in front of the basic verb "no" as shown in FIG. 8B, and an extended stem form of the compound verb "No", In other words, the basic, irregular, continuous, present, connected, past, future, terminating and noun forms and their respective surviving forms are automatically combined by combining "A" in front of each form of the extended stem of the basic verb "no". Can be created with In this way, by expanding the stem of a single word, which is a basic term, and extending stems such as conjoined words and compound words by using the extended stem of a single word, it is possible to add the actually used forms of verbs and adjectives to the word semantic dictionary DB. .

The operation by the grammar checking method according to the embodiments described above may be implemented at least partially as a computer program and recorded in a computer-readable recording medium. A recording medium in which a program for implementing an operation according to the method according to the embodiments is recorded and the computer-readable recording medium includes all kinds of recording devices in which data that can be read by a computer is stored. Examples of computer-readable recording media include ROM, RAM, CD-ROM, magnetic tape, floppy disk, and optical data storage devices. In addition, the computer-readable recording medium may be distributed over a computer system connected through a network, and computer-readable codes may be stored and executed in a distributed manner. In addition, functional programs, codes, and code segments for implementing this embodiment may be easily understood by those skilled in the art to which this embodiment belongs.

The present invention described above has been described with reference to the embodiments shown in the drawings, but these are merely exemplary, and those of ordinary skill in the art will understand that various modifications and variations of the embodiments are possible therefrom. However, such modifications should be considered to be within the technical protection scope of the present invention. Therefore, the true technical scope of the present invention should be determined by the technical spirit of the appended claims.

Claims (11)

As a grammar checking system,
A storage module configured to store a word semantic dictionary database including a plurality of words and attribute information corresponding to each of the plurality of words, and processing rule information defined using the attribute information;
An input module configured to extract one or more sentences from input data, and to separate one or more words or phrases from the extracted one or more sentences based on the word semantic dictionary database;
A grammar checking module configured to detect a grammar error by applying a grammar processing rule defined using the attribute information to one or more separated words or phrases; And
Comprising an output module configured to provide information on an error detected by the grammar checking module,
The attribute information is defined step by step including a first attribute indicating a component of a word, a second attribute indicating a part of speech of a word, and a third attribute indicating a semantic property classification of the word, and the third attribute is a component and a part of speech. Is defining a group of words having the same applied grammar among the same words, and at least one third attribute corresponding to a lower level of the second attribute is defined for one of the second attribute,
The grammar checking system is further configured to determine the first attribute, the second attribute, and the third attribute corresponding to the separated one or more words or phrases,
The processing rule information,
Including one or more of a spacing rule, a search use rule, a ending rule, and a word combination rule defined using the third attribute of a word,
A grammar checking system in which grammar rules for converting words or characters are converted into a language readable by a computing device, displayed to a user, and defined in the form of a processing function that can be modified by the user.
delete The method of claim 1,
The processing function is a grammar checking system that is executed on condition of at least one of a document type, a sentence position, and an adjacent character.
The method of claim 1,
Grammar inspection system further comprising a term management module configured to generate a compound verb with a conjugated form defined by combining a basic term with a pre-defined conjugated form with other words to expand the stem, and store the compound term in the word semantic dictionary database .
The method of claim 1,
A typo generator configured to generate a predicted typo by substituting a part of a word included in the word semantic dictionary database by using a preset character or group of words; And
A grammar checking system further comprising a typo checking module comprising a typo checking unit configured to perform a typo checking on the separated one or more words or words based on the expected typo generated by the typo generating unit.
Storing a word semantic dictionary database including a plurality of words and attribute information corresponding to each of the plurality of words in a grammar checking system;
Storing processing rule information defined by using the attribute information in the grammar checking system;
Extracting, by the grammar checking system, one or more sentences from input data;
Separating, by the grammar checking system, one or more words or phrases from the extracted one or more sentences based on the word meaning dictionary database;
Detecting, by the grammar checking system, a grammar error by applying a grammar processing rule defined using the attribute information to one or more separated words or phrases; And
Including the step of the grammar checking system, providing information on the detected error,
The attribute information is defined step by step including a first attribute indicating a component of a word, a second attribute indicating a part of speech of a word, and a third attribute indicating a semantic property classification of the word, and the third attribute is a component and a part of speech. Is defining a group of words having the same applied grammar among the same words, and a plurality of the third attributes corresponding to a lower level of the second attribute are defined for one of the second attributes,
Before the step of detecting the grammar error, the grammar checking system further comprises determining the first attribute, the second attribute and the third attribute corresponding to the separated one or more words or words,
The processing rule information,
Including one or more of a spacing rule, a survey use rule, a ending rule, and a word combination rule defined using the third attribute of a word,
A grammar checking method in which a grammar rule for converting a word or letter is converted into a language readable by a computing device and displayed to a user, and is defined in the form of a processing function that can be modified by the user.
delete The method of claim 6,
The processing function is a grammar checking method that is executed on condition of one or more of a document type, a sentence position, and an adjacent character.
The method of claim 6,
Generating, by the grammar test system, a compound verb having a conjugated form defined by combining a basic term with a pre-defined conjugation form with another word to expand a stem; And
The grammar checking method further comprises the step of storing, by the grammar checking system, the compound verb in the word meaning dictionary database.
The method of claim 6,
Generating, by the grammar checking system, a predicted typo by substituting a part of words included in the word semantic dictionary database by using a preset letter or group of words; And
And performing, by the grammar checking system, a typo check for the separated one or more words or words based on the generated predicted typo.
A computer program stored in a computer-readable recording medium to execute the grammar checking method according to any one of claims 6 and 8 to 10 in combination with hardware.

Images