KR20210055175A - Apparatus and method for measuring difficulty level of chinese character using regression analysis - Google Patents

Abstract

Disclosed are a Chinese character difficulty level measuring device using regression analysis and a method thereof. According to an embodiment of the present invention, the Chinese character difficulty level measuring device using regression analysis comprises the following steps of: preparing a first data set about a Chinese character usage frequency from sample text data about each of predetermined Chinese characters; preparing a second data set for a Korean character meaning frequency from the sample text data about each of the predetermined Chinese characters; and measuring a difficulty level of each of the predetermined Chinese characters through multiple regression analysis based on the first data set, the second data set, and stroke count information of each of the predetermined Chinese characters.

Description

A device and method for measuring Chinese character difficulty using regression analysis {APPARATUS AND METHOD FOR MEASURING DIFFICULTY LEVEL OF CHINESE CHARACTER USING REGRESSION ANALYSIS}

The present application relates to an apparatus and method for measuring difficulty of Chinese characters using regression analysis.

Currently, there are nine institutions that host the nationally recognized Chinese character test, including the Korean Language Literature Association, the Chinese Character Education Promotion Association, and the Korean Language Proficiency Test Association. However, different Chinese characters are assigned for each series of places that host the Chinese character test, and there are cases where the criteria for assignment are unclear or have not been disclosed at all. In the case of such a series of Chinese characters, they are usually classified from level 8 to level 1, and the more difficult the Chinese characters are included in level 1.

Specifically, looking at the Chinese characters for each grade assigned by the Korean Literature, the level of Chinese character allocation at level 8, which is the lowest level of difficulty for those who are new to Chinese characters, is called'a series for motivating Chinese character learning (50 common Chinese characters). It is only specified, and if you look at the assigned kanji, you can easily find cases that contain kanji that are somewhat difficult for beginners to learn.

Also, looking at the kanji test table provided by the Korea Chamber of Commerce and Industry, Chinese characters are allocated for elementary school students in grades 9 to 5, and in the case of grade 9, ‘Korean-Korean mixed sentences using ordinary Chinese characters at the level of the first grade of elementary school are used. It is written that the level of Chinese character ability to read and understand to a certain extent', but it is difficult to confirm how often the corresponding Chinese characters are used in the first grade of elementary school.

Chinese characters are slogans, and each character, that is, each character contains its meaning and sound, and by this characteristic, it is essential to recognize the meaning of each character in order to read a Chinese character and understand its meaning accurately. In addition, there are differences between expressions used as meanings (meanings) of Chinese characters and Korean words commonly used in real life. For example, the meaning of'metal gold (金)' in the 8th class assigned Chinese characters selected by the Korea Chamber of Commerce and Industry. The meaning of Soo rarely appears in elementary school textbooks, and even in the case of'author city','author' has the same meaning as the mayor and the street, but rarely appears in elementary school textbooks.

As such, it is difficult to say that the difficulty of Chinese characters for each series provided by the institution hosting the conventional nationally recognized Chinese character test is evaluated on an objective basis, and there is no technology that can evaluate the difficulty of Chinese characters on the basis of objective indicators. .

The technology behind the present application is disclosed in Korean Patent Publication No. 10-0779022.

The present application is to solve the problems of the prior art described above, using multiple linear regression analysis based on big data (data set) collected and extracted from sample text data that can reflect the learning level of the subject, such as elementary school textbooks. An object of the present invention is to provide an apparatus and method for measuring difficulty of Chinese characters using regression analysis capable of objectively providing the difficulty level for each predetermined Chinese character.

However, the technical problem to be achieved by the embodiments of the present application is not limited to the technical problems as described above, and other technical problems may exist.

As a technical means for achieving the above-described technical problem, the method of measuring Chinese character difficulty using regression analysis according to an embodiment of the present application includes, for each predetermined Chinese character, a first data set for the frequency of using Chinese characters from sample text data. Preparing, for each of the predetermined Chinese characters, preparing a second data set for the Hangul meaning frequency from the sample text data, and the number of strokes of each of the first data set, the second data set, and the predetermined Chinese characters It may include the step of measuring the difficulty of each of the predetermined Chinese characters through multiple regression analysis on the basis of the information.

In addition, the preparing of the first data set may include collecting sentences appearing in the sample text data, stating a Chinese character notation corresponding to a Hangul part corresponding to a Chinese character word in each of the sentences, and the Chinese character notation. It may include the step of counting the frequency of use of the Chinese characters by mapping each of the sentences in which a is added and each of the predetermined Chinese characters.

In addition, the preparing of the second data set includes tokenizing the collected sentences, rooting words segmented through the tokenization, and rooting converted through the rooting and the predetermined Chinese character. It may include the step of counting the Hangul semantic frequency by mapping each.

In addition, in the counting the Hangul semantic frequency, the Hangul semantic frequency may be counted based on the original word included in the converted root.

In addition, the counting the Hangul semantic frequency may perform the mapping in consideration of the synonyms of each of the predetermined Chinese characters.

In addition, the sample text data may include elementary textbooks for each grade, and the first data set and the second data set may be divided and prepared for each grade.

In addition, in the step of measuring the difficulty level of each of the predetermined Chinese characters, when receiving information about an appropriate learning time for each of the predetermined Chinese characters, the difficulty level may be measured by further considering the appropriate learning time information.

In addition, the method of measuring the difficulty of Chinese characters using regression analysis according to an exemplary embodiment of the present disclosure may include measuring the difficulty of words composed of Chinese characters based on the difficulty of each of the predetermined Chinese characters.

In addition, in measuring the difficulty of the word, the average value of the difficulty measured for each Chinese character constituting the word may be determined as the difficulty of the word.

In addition, the method of measuring Chinese character difficulty using regression analysis according to an embodiment of the present application includes some of the predetermined Chinese characters corresponding to a predetermined grade based on the difficulty level of each of the predetermined Chinese characters and the frequency of use of the Chinese characters. It may include the step of classifying as.

In addition, the step of classifying Chinese characters for learning may include filtering Chinese characters whose utilization frequency of the Chinese characters in all grades is equal to or higher than a preset threshold frequency among the predetermined Chinese characters based on the first data set, and among the filtered Chinese characters. Based on the first data set prepared for each grade, the frequency of using the Chinese characters in the corresponding grade is greater than or equal to a preset threshold frequency for each grade, and the learning of Chinese characters having a difficulty level that falls within a preset difficulty range for the corresponding grade. It may include the step of selecting a target Chinese character.

On the other hand, the Chinese character difficulty measuring apparatus using the regression analysis according to an embodiment of the present application, for each predetermined Chinese character, a Chinese character utilization frequency acquisition unit for preparing a first data set for the Chinese character utilization frequency from sample text data, the predetermined For each of the Chinese characters, a Hangul semantic frequency acquisition unit for preparing a second data set for Hangul semantic frequencies from the sample text data, and the first data set, the second data set, and the stroke count information of each of the predetermined Chinese characters. Based on the multiple regression analysis, it may include a difficulty analysis unit that measures the difficulty level of each of the predetermined Chinese characters.

In addition, the sample text data includes elementary textbooks for each grade, the Chinese character utilization frequency acquisition unit prepares the first data set divided by grade, and the Hangul semantic frequency acquisition unit prepares the second data set classified by grade. can do.

In addition, when receiving information about an appropriate learning time for each of the predetermined Chinese characters, the difficulty analysis unit may measure the difficulty by further considering the appropriate learning time information.

In addition, the apparatus for measuring Chinese character difficulty using regression analysis according to an embodiment of the present application includes some of the predetermined Chinese characters corresponding to a predetermined grade based on the difficulty level of each of the predetermined Chinese characters and the frequency of use of the Chinese characters. It may include a learning target Chinese character determination unit classified as.

The above-described problem solving means are merely exemplary and should not be construed as limiting the present application. In addition to the above-described exemplary embodiments, additional embodiments may exist in the drawings and detailed description of the invention.

According to the above-described problem solving means of the present application, the difficulty of each predetermined Chinese character is determined using multiple linear regression analysis based on the big data collected and extracted from sample text data that can reflect the learning level of the subject, such as elementary school textbooks. There is an effect of providing an apparatus and method for measuring difficulty of Chinese characters using regression analysis that can be provided objectively.

According to the above-described problem solving means of the present application, it is possible to provide an objective basis for calculating the difficulty of Chinese characters.

According to the above-described problem solving means of the present application, it is possible to objectively measure the degree of difficulty of each Chinese character, so that a predetermined Chinese character can be classified for each grade, or a list of Chinese characters for each series in consideration of the level of difficulty can be provided.

According to the above-described problem solving means of the present application, a greater motivation for learning Chinese characters can be provided to a learner by providing a Chinese character to be learned in consideration of vocabulary and words actually used in a textbook.

According to the above-described problem solving means of the present application, it is possible to provide a customized learning service in consideration of the level of the learner.

However, the effect obtainable in the present application is not limited to the above-described effects, and other effects may exist.

1 is a schematic configuration diagram of a Chinese character difficulty measuring system including a Chinese character difficulty measuring apparatus using regression analysis according to an embodiment of the present application.
2 is a conceptual diagram illustrating a process of obtaining a Chinese character utilization frequency by a Chinese character difficulty measuring apparatus using regression analysis according to an embodiment of the present application.
3 is a diagram showing a sentence in which a collected sentence and a Chinese character notation corresponding to a Hangul part corresponding to a Chinese character word in the corresponding sentence are also included.
4 is a conceptual diagram illustrating a process of obtaining a Hangul semantic frequency by a Chinese character difficulty measuring apparatus using regression analysis according to an embodiment of the present application.
5 is a chart showing collected sentences, tokenization results and rootization results for the sentences.
6 is a chart showing the meaning, sound, the first data set, the second data set, and the number of strokes of each of the predetermined Chinese characters.
FIG. 7 is a diagram illustrating by exemplarily selecting some of the characteristics to be subjected to multiple regression among characteristics including a first data set, a second data set, and stroke number information according to a predetermined criterion.
8 is a diagram exemplarily showing weight values and bias values for each characteristic in a Chinese character difficulty prediction equation completed through multiple regression analysis.
9 is a chart showing a result of measuring the difficulty of each predetermined Chinese character by a Chinese character difficulty measuring apparatus using a regression analysis according to an embodiment of the present application.
10 is a chart showing the number of learning target kanji selected for each grade and a preset difficulty range for each grade for classifying kanji for learning by grade.
11 is an experimental example linked to a device for measuring Chinese character difficulty using a regression analysis according to an embodiment of the present application, and is a chart showing the result of classifying Chinese characters to be studied for a first grader of an elementary school.
12 is a schematic configuration diagram of a Chinese character difficulty measuring apparatus using regression analysis according to an embodiment of the present application.
13 is a flowchart illustrating a method of measuring Chinese character difficulty using regression analysis according to an exemplary embodiment of the present application.

Hereinafter, embodiments of the present application will be described in detail with reference to the accompanying drawings so that those of ordinary skill in the art may easily implement the present application. However, the present application may be implemented in various different forms and is not limited to the embodiments described herein. In the drawings, parts not related to the description are omitted in order to clearly describe the present application, and similar reference numerals are attached to similar parts throughout the specification.

Throughout this specification, when a part is said to be "connected" with another part, it is not only a case that it is "directly connected", but also "electrically connected" or "indirectly connected" with another element interposed therebetween. "Including the case.

Throughout the present specification, when a member is positioned "on", "upper", "upper", "under", "lower", and "lower" of another member, this means that a member is located on another member. This includes not only the case where they are in contact but also the case where another member exists between the two members.

In the entire specification of the present application, when a certain part "includes" a certain component, it means that other components may be further included rather than excluding other components unless specifically stated to the contrary.

1 is a schematic configuration diagram of a Chinese character difficulty measuring system including a Chinese character difficulty measuring apparatus using regression analysis according to an embodiment of the present application.

Referring to FIG. 1, the Chinese character difficulty measuring system 10 according to an embodiment of the present application considers the frequency of using Chinese characters and the frequency of Korean meaning extracted from the sample text data 1 through analysis of the sample text data 1 Thus, it may mean a system that measures (evaluates) the difficulty level of each Chinese character included in a predetermined Chinese character (2) or selects some Chinese characters from the predetermined Chinese characters (2) as learning target Chinese characters.

In the description of the embodiment of the present application, the predetermined Chinese character (2) may mean 1800 common Chinese characters designated by the Ministry of Education, but is not limited thereto. For example, the predetermined kanji (2) may be a kanji for each grade in a kanji test provided by an institution such as the Korean Literature Society, the Chinese Character Education Promotion Association, and the Korean Examination Association. As another example, the predetermined Chinese character 2 may be determined as an arbitrary set of Chinese characters including a plurality of Chinese characters.

Illustratively for better understanding, when the predetermined Chinese character (2) is 1800 common Chinese characters designated by the Ministry of Education, the Chinese character difficulty measuring device 100 using a regression analysis according to an embodiment of the present application (hereinafter,'Chinese character difficulty measuring device 100 )'.) can be understood as a device that evaluates and provides the difficulty level of each of 1800 Chinese characters included in 1800 common Chinese characters designated by the Ministry of Education based on big data obtained by analyzing sample text data (1). .

In addition, in the description of the embodiment of the present application, the sample text data (1) may mean an elementary textbook for each grade, but is not limited thereto. For example, the sample text data (1) may include arbitrary books according to the purpose, type, etc. of the Chinese character difficulty to be calculated through the Chinese character difficulty measuring device 100 of the present application, or may be determined as a secondary school textbook, a high school textbook, etc. have.

For better understanding, if the sample text data (1) is an elementary textbook for each grade, the difficulty level for each of the predetermined Chinese characters (2) may be measured in consideration of the educational curriculum of general elementary school students. When the data (1) is a middle school textbook or a high school textbook, the difficulty level for each of the predetermined Chinese characters (2) may be measured in consideration of the educational curriculum of a general middle school student or high school student.

In other words, a predetermined Chinese character (2) in the present application is an evaluation target for measuring the difficulty level, and the sample text data (1) may mean learning data that is a standard for measuring the difficulty level for a predetermined Chinese character (2). . However, for convenience of explanation, the following description of the embodiments of the present application is based on the fact that the predetermined Chinese character (2) is 1800 common Chinese characters designated by the Ministry of Education, and the sample text data (1) is selected as an elementary textbook for each grade. do.

The Chinese character difficulty measuring apparatus 100 may prepare a first data set (A) for the frequency of using Chinese characters from sample text data (1) for each of the predetermined Chinese characters (2).

2 is a conceptual diagram illustrating a process of obtaining a Chinese character utilization frequency by a Chinese character difficulty measuring apparatus using regression analysis according to an embodiment of the present application.

Referring to FIG. 2, the Chinese character difficulty measuring apparatus 100 collects sentences appearing in sample text data 1, and performs a Chinese character representation corresponding to a Hangul part corresponding to a Chinese character word in each of the collected sentences. Can be marked sentence.

3 is a diagram showing a sentence in which a collected sentence and a Chinese character notation corresponding to a Hangul part corresponding to a Chinese character word in the corresponding sentence are also included.

Referring to FIG. 3, the Chinese character difficulty measuring apparatus 100 may collect, for example, a sentence "Let's say hello to the people we met at school" among sentences appearing in the sample text data 1, and the collected sentences. In, the Chinese characters corresponding to the Hangul part (school, personnel) corresponding to the Chinese word can be written together. Referring to FIG. 3 with respect to the staging method, as an example, the Chinese character difficulty measuring apparatus 100 may adopt a method in which parentheses are displayed in succession with the Hangul part and a Chinese character corresponding to the Hangul part is written in the parentheses. It is not limited to this.

According to the exemplary embodiment of the present application, the Chinese character notation corresponding to the Hangul part may be performed using a Hangul Hanja automatic conversion system such as Utagger.

In addition, the Chinese character difficulty measuring apparatus 100 may count the frequency of using Chinese characters by mapping each of a marked sentence with a Chinese character and a predetermined Chinese character 2.

According to an exemplary embodiment of the present application, the frequency of using Chinese characters for each of the predetermined Chinese characters 2 may be counting the number of times the corresponding Chinese characters appear in a sentence set in which Chinese characters are notation in the collected sentences.

In addition, according to an embodiment of the present application, the Chinese character difficulty measuring apparatus 100 examines and corrects whether the Chinese character utilization frequency for each of the obtained predetermined Chinese characters (2) is correctly counted, so that the first data set (A) Can be confirmed.

For example, the Chinese character difficulty measuring device 100 checks whether the Chinese character is correctly determined to match the meaning of the corresponding Korean part by searching for a marked sentence in which the Chinese character is written, and if an inappropriate Chinese character is written, the corresponding Chinese character is displayed. You can modify your weapon.

In addition, the Chinese character difficulty measuring apparatus 100 may prepare a second data set B for Hangul semantic frequency from the sample text data 1 for each of the predetermined Chinese characters 2.

4 is a conceptual diagram illustrating a process of obtaining a Hangul semantic frequency by a Chinese character difficulty measuring apparatus using regression analysis according to an embodiment of the present application.

Referring to FIG. 4, the Chinese character difficulty measuring apparatus 100 may collect sentences appearing in sample text data and tokenize each of the collected sentences. In addition, the Chinese character difficulty measuring apparatus 100 may stemming each word in a segmented sentence through tokenization.

Tokenizing is a term that refers to the work of dividing a given corpus in units of tokens. Segmenting sentences using words as token units, removing punctuation marks, and making sentences based on spaces. It can refer to tasks such as segmenting.

Stemming may be referred to differently as stem extraction, etc., and may refer to the work of standardizing by removing or changing the ending around the root.

5 is a chart showing collected sentences, tokenization results and rootization results for the sentences.

Referring to FIG. 5, the Chinese character difficulty measuring apparatus 100 may collect, for example, a sentence “Let’s say hello to the people we met at school” among sentences appearing in the sample text data 1, and the collected sentences. The sentence can be segmented in units of words through tokenizing for. Referring to FIG. 5, the results segmented by tokenization are'School/Noun','At/Josa','Meet/Noun','People/Noun','Suffix','Gwa/Josa', It can be expressed as'Greetings/Noun','Sun/Verb','Let's see/Verb', and'./Punctuation'.

According to an exemplary embodiment of the present disclosure, the Chinese character difficulty measuring apparatus 100 may obtain part of speech information of each segmented word through tokenization. For example, referring to FIG. 5, part of speech information of each segmented word may be displayed in the form of'/ part of speech' in the segmented word. Among parts of speech notation, Noun is a noun, Josa is a survey, Suffix is a suffix, Verb may refer to verbs and punctuation may refer to punctuation marks. However, the display of parts of speech is not limited to this.

In addition, referring to FIG. 5, the result of rooting the segmented words through tokenization (transformed root) is'School/Noun','At/Josa','Met/Noun','People/Noun', It can be expressed as'Deul/Suffix','And/Josa','Greeting/Noun','Hada/Verb','Boda/Verb', and'./Punctuation', respectively.

Specifically, if the tokenization result and the rooting result are compared,'Let's see' may be converted into'look' and'sun' may be converted to'hada' according to the rooting.

In addition, according to an embodiment of the present application, the Chinese character difficulty measuring apparatus 100 may perform rooting using a Hangul morpheme analyzer such as Open-Korean-Txt (OKT).

The Chinese character difficulty measuring apparatus 100 may count the Hangul meaning frequency by mapping each of the roots converted through rooting and a predetermined Chinese character 2. In addition, according to an exemplary embodiment of the present disclosure, when counting the Hangul meaning frequency, the Chinese character difficulty measuring apparatus 100 may count the Hangul meaning frequency based on the original word included in the converted root. For example, when the original word'look through' is converted into the main word'look through' and the supplementary word'bather' by rooting, the Chinese character difficulty measuring device 100'looks through', which corresponds to the main term representing the actual meaning. The Hangul semantic frequency of at least one of the predetermined Chinese characters (2) corresponding to 'is counted, but the Hangul semantic frequency of the Chinese characters corresponding to'more' corresponding to the auxiliary term indicating the grammatical meaning is not counted. I can.

In addition, according to an exemplary embodiment of the present disclosure, the apparatus 100 for measuring difficulty of Chinese characters may count the frequency of Hangul meanings by performing mapping in consideration of the synonyms of each of the predetermined Chinese characters 2. For example, in relation to the transformed root'Boda/Verb' (however, it is assumed here that'Boda' is used as this term). The frequency of Hangul meanings of Chinese characters such as) may be counted.

In addition, according to an embodiment of the present application, the Chinese character difficulty measuring apparatus 100 may count the frequency of Hangul meaning in consideration of the context in the sentence in which the converted root appears. For example, with respect to the transformed root'eye/Noun', the Chinese character difficulty measuring device 100 considers the context in the sentence in which the transformed root appears, and is a sensory organ that can see an object by being stimulated by light. It distinguishes whether it means'snow' or'snow', which is a crystal of ice falling on the ground by freezing when water vapor in the atmosphere meets a cold energy, and the Hangul meaning for a predetermined Chinese character that has a meaning consistent with the analyzed meaning. You can count the frequency. In other words, the Chinese character difficulty measuring apparatus 100 may count the frequency of Hangul meanings in consideration of the homophones.

In addition, according to an embodiment of the present application, the Chinese character difficulty measuring apparatus 100 examines and corrects whether the Hangul semantic frequency for each of the obtained predetermined Chinese characters (2) is correctly counted, so that the second data set (B) Can be confirmed.

6 is a chart showing the meaning, sound, the first data set, the second data set, and the number of strokes of each of the predetermined Chinese characters.

Referring to FIG. 6, information on each of a first data set (A), a second data set (B), and a predetermined Chinese character (2) may be obtained in the form of a table.

Specifically, an identification number (idhanja) may be assigned to each of the predetermined Chinese characters (2) (for example, 0 to 1799 in the case of 1800 common Chinese characters designated by the Ministry of Education), and the Chinese characters (hanja) for each Chinese character, um (sound), meaning, and number of strokes ( St , C) can be obtained together.

In addition, according to an embodiment of the present application, when the sample text data 1 is an elementary textbook for each grade, the first data set A and the second data set B may be divided and prepared (obtained) for each grade. . According to an embodiment of the present application, the sample text data (1) is a Korean language textbook from the first grade to the sixth grade of elementary school, the right life of the first and second graders of elementary school, a wise life, a fun life textbook, and from the third to sixth grade of elementary school. May include social textbooks. However, it is not limited thereto, and other sample text data 1 may be used according to changes in the elementary curriculum.

For example, referring to FIG. 6, FHU is the frequency of using Chinese characters for each predetermined Chinese character (2) in the entire grade (1st to 6th grade) elementary textbook, and FHU_1 to FHU_6 are in elementary textbooks for each grade. It may be the frequency of using the Chinese characters for each of the predetermined Chinese characters (2). That is, FHU_1 may be the frequency of using Chinese characters in first grade elementary textbooks, and FHU_6 may be the frequency of using Chinese characters in sixth grade elementary textbooks.

Likewise, referring to FIG. 6, FWS is the Hangul meaning frequency for each predetermined Chinese character (2) in the entire grade (1st to 6th grade) elementary textbook, and FWS_1 to FWS_6 are predetermined in elementary textbooks for each grade. It may be the frequency of Hangul meaning for each of the Chinese characters (2). That is, FWS_1 may be the frequency of Hangul meaning in the first grade elementary textbook, and FWS_6 may be the Hangul semantic frequency in the sixth grade elementary textbook.

In addition, the Chinese character difficulty measuring apparatus 100 is based on the first data set (A), the second data set (B), and the stroke number information (C) of each of the predetermined Chinese characters (2), multiple linear regression analysis (Multiple Linear Regression). It is possible to measure the difficulty of each of the predetermined Chinese characters (2). Referring to FIG. 6, the feature for applying multiple regression analysis is 1 stroke count information ( St , C), 6 first data sets (FHU_n ) collected separately for each grade, and a first data set for each grade (FHU). ) One, six second data sets ( FWS_n ) collected separately for each grade, and one second data set ( FWS ) for each grade may be combined to be 15, but is not limited thereto. As another example, as will be described later, the Chinese character difficulty measuring apparatus 100 may apply multiple regression analysis by selecting some of the characteristics (eg, 15) obtained according to a predetermined criterion.

Hereinafter, with reference to FIGS. 7 and 8, a process in which the Chinese character difficulty measuring apparatus 100 measures the difficulty level for each of the predetermined Chinese characters 2 through multiple regression analysis will be described in detail.

According to an embodiment of the present application, the Chinese character difficulty measuring apparatus 100 may select some of features including a first data set (A), a second data set (B), and stroke number information (C). have. For example, the Chinese character difficulty measuring apparatus 100 may select some of the acquired characteristics using a stepwise regression analysis. Specifically, the Chinese character difficulty measuring apparatus 100 repeats the process of verifying the learned multiple regression analysis model in a state in which some of the features are included or excluded using Stepwise Regression Analysis to select a feature. I can.

The process of selecting some of the features including the first data set (A), the second data set (B), and the number of strokes information (C) is illustratively the total Chinese character utilization frequency ( FHU ) at a critical level. It may be repeatedly performed in a manner excluding some Chinese characters among the predetermined Chinese characters (2) written below. According to an exemplary embodiment of the present disclosure, a threshold level that is a criterion for excluding some Chinese characters may be set differently for each repeated trial.

According to an embodiment of the present application, verifying the multiple regression model in the state of including or excluding some of the features is a coefficient of determination, which is a measure of the degree to which the estimated linear model fits a given data in each model. It may be verified by calculating (coefficient of determination).

In addition, according to an embodiment of the present disclosure, verifying a multiple regression analysis model in a state in which some of the features are included or excluded may be verified by calculating a root mean squared error (RMSE) in each model. . Specifically, the RMSE value for each model may be calculated by Equation 1 below.

[Equation 1]

FIG. 7 is a diagram illustrating by exemplarily selecting some of the characteristics to be subjected to multiple regression among characteristics including a first data set, a second data set, and stroke number information according to a predetermined criterion.

Referring to FIG. 7, as a result of repeatedly performing verification of the multiple regression analysis model in the state of including or excluding some of the characteristics, in consideration of each coefficient of determination, stroke number information ( St , C ), Hangul semantic frequency for 3rd grade textbooks ( FWS_3 ), Hangul semantic frequency for 5th grade textbooks ( FWS_5 ), Chinese character usage frequency for 1st grade textbooks ( FHU_1 ), Chinese character usage frequency for 2nd grade textbooks ( FHU_2 ), and The frequency of using Chinese characters ( FHU_4 ) for 4th grade textbooks can be selected as a characteristic. Referring to FIG. 7, it can be seen that in the case of a multiple regression model including the above-described characteristics, the coefficient of determination value is 0.852, which is the best result compared to other trials.

In addition, referring to FIG. 7, it can be seen that when a Chinese character having a total Chinese character utilization frequency (FHU ) of 2 or more among the predetermined Chinese characters (2) is used, the coefficient of determination shows the best result. Therefore, the Chinese character difficulty measuring apparatus 100 has the total Chinese character utilization frequency of the first data set (A), the second data set (B), and the stroke number information (C) for each of the prepared predetermined Chinese characters (2) at a critical level ( Multiple regression can be performed using selected characteristics (St , FWS_3 , FWS_5 , FHU_1 , FHU_2 , FHU_4 ) of Chinese characters with FHU ≥2) or higher.

However, the characteristics selected by the above-described process and the threshold level for the total Chinese character utilization frequency should be understood as exemplary matters that may change depending on which sample text data (1) or a predetermined Chinese character (2) is set, It should not be construed to limit the technical idea of the present application.

According to an embodiment of the present application, the Chinese character difficulty measuring apparatus 100 is a Chinese character filtered according to a threshold level determined for a selected characteristic and a Chinese character utilization frequency among a predetermined Chinese character (2) (e.g., Chinese characters having an FHU of 2 or more). ), a weight value and a bias value for each feature may be calculated by applying multiple regression analysis on the basis of the selected feature to complete the Chinese character difficulty prediction equation.

8 is a diagram exemplarily showing a weight value and a bias value for each characteristic in a Chinese character difficulty prediction equation completed through multiple regression analysis.

Referring to FIG. 8, the Chinese character difficulty measuring apparatus 100 sets the weight for the normalized stroke number information (NSt, C) to 0.6164 and the weight for the Hangul semantic frequency (NFWS_3) for the normalized third grade textbook to -1.1085. , The weight of the Hangul semantic frequency (NFWS_5) for the normalized fifth grade textbook is -0.2659, the weight for the Chinese character utilization frequency (NFHU_1) for the normalized first grade textbook is -0.2659, and the normalized second grade textbook is The Chinese character difficulty prediction equation can be completed by determining the weight for the frequency of using Chinese characters (NFHU_2) as 0.2542, the weight for the frequency of using Chinese characters (NFHU_4) for the normalized 4th grade textbook as -0.3491, and the Bias value as 0.4096. . The completed Chinese character difficulty prediction equation is expressed as Equation 2 below.

[Equation 2]

Here, C is a variable representing each Chinese character included in a predetermined Chinese character (2), and D may mean a difficulty prediction function (expression). That is, when the Chinese character difficulty measuring apparatus 100 completes the Chinese character difficulty prediction equation in the form of Equation 2, the six characteristic values for each Chinese character of a predetermined Chinese character (2) are substituted into the equation to obtain the difficulty value for the corresponding Chinese character. can do.

In addition, according to an embodiment of the present application, normalized data for each of the selected characteristics is divided into training data (Train) and test data (Test set), and normalized using MinMaxScaler. It can be obtained through the process. For example, the training data (Train) may be selected as 80% of the total data, and the test data (Test set) may be selected as 20% of the total data.

In addition, according to an embodiment of the present application, when receiving appropriate learning timing information for each of the predetermined Chinese characters (2), the Chinese character difficulty measuring device 100 may measure the difficulty by further considering the received appropriate learning timing information. I can.

According to an embodiment of the present application, the appropriate learning time information may be applied to the Chinese character difficulty measuring apparatus 100 from a separate database or server that stores a questionnaire result for an appropriate learning time for each predetermined Chinese character (2). have.

According to the exemplary embodiment of the present application, the appropriate learning time information may be generated in a form in which the appropriate learning time for each predetermined Chinese character (2) is determined as 12 values from the 1st semester of the 1st grade to the 2nd semester of the 6th grade. For example, when it is determined that the appropriate learning time for any one Chinese character is '3rd grade 1st semester', the information on the appropriate learning time for the corresponding Chinese character may have a value of '5'.

The Chinese character difficulty measuring apparatus 100 uses the average of the values corresponding to 95% (2σ) of the distribution by considering the standard deviation of a plurality of appropriate learning time information for each of the received predetermined Chinese characters (2). The learning timing information can be confirmed. As such, the apparatus 100 for measuring Chinese character difficulty may be implemented to correct a weight value and a Bias value in a Chinese character difficulty prediction equation based on the appropriate learning time information when the predetermined Chinese character has appropriate learning time information.

9 is a chart showing a result of measuring the difficulty of each predetermined Chinese character by a Chinese character difficulty measuring apparatus using a regression analysis according to an embodiment of the present application.

Referring to FIG. 9, the difficulty level of each Chinese character provided by the Chinese character difficulty measuring apparatus 100 may be measured in DL. Here, it can be understood that the higher the value of DL, the higher the difficulty.

In addition, the Chinese character difficulty measuring apparatus 100 may measure the difficulty of a word composed of Chinese characters based on the measured difficulty of each of the predetermined Chinese characters 2. According to an exemplary embodiment of the present disclosure, the Chinese character difficulty measuring apparatus 100 may determine an average value of the measured difficulty for each Chinese character constituting a word for which the difficulty level is to be measured, as the difficulty of the word. For example, in the case of the word'firefighter', which is a word composed of Chinese characters, the difficulty level of the Chinese character'small' measured by the Chinese character difficulty measuring device 100 through multiple regression analysis, and the Chinese character'bang (防) You can measure (evaluate) the average of the difficulty level for')' and the difficulty level for the Chinese character'Bang (官)' as the difficulty level of the word.

Furthermore, the Chinese character difficulty measuring device 100 can measure the overall difficulty of text including Chinese characters (original Chinese data, documents in which Korean and Chinese are mixed, etc.) based on the measured difficulty of each of the predetermined Chinese characters (2). have.

In addition, according to an embodiment of the present application, the Chinese character difficulty measuring apparatus 100 includes a text including Hangul (for example, the entire Chinese character was written in Korean, but the Chinese character When the corresponding Hangul part is included), the level of difficulty of the kanji of the part corresponding to the kanji word can be collected and the readability of the corresponding text can be evaluated. For example, the readability of the text may be evaluated in a manner that is selected as any one of a plurality of preset grades.

In addition, the Chinese character difficulty measuring apparatus 100 may classify some of the predetermined Chinese characters 2 into a learning target Chinese character corresponding to a predetermined grade based on the measured difficulty of each of the predetermined Chinese characters 2 and the frequency of use of the Chinese characters. have.

Specifically, the Chinese character difficulty measuring apparatus 100 filters Chinese characters whose use frequency (FHU ) of Chinese characters in all grades is greater than or equal to a preset threshold frequency among predetermined Chinese characters (2) based on the first data set (A). Based on the first data set (A, FHU_1 to FHU_6 ) prepared for each grade, the frequency of using Chinese characters in the grade is greater than or equal to the preset threshold frequency for each grade, and has a difficulty level that falls within the preset difficulty range for the corresponding grade. Can be selected as the target Chinese character for the relevant grade.

According to an embodiment of the present application, a preset threshold frequency for the Chinese character utilization frequency (FHU) in the entire grade may be 5. That is, the Chinese character difficulty measuring apparatus 100 may firstly select (filter) Chinese characters having a frequency of using Chinese characters of 5 or more in the textbook for all grades.

In addition, according to an exemplary embodiment of the present application, a preset threshold frequency for each grade level for the Chinese character utilization frequency (FHU_n) in a corresponding grade may be 1. That is, the Chinese character difficulty measuring apparatus 100 may secondaryly select (filter) Chinese characters having a frequency of using one or more Chinese characters in a textbook for a corresponding grade.

Subsequently, the Chinese character difficulty measuring apparatus 100 uses a Chinese character having a difficulty level that falls within a preset difficulty range for a corresponding grade from among the first selected (filtered) and second selected (filtered) Chinese characters as a learning target Chinese character for the corresponding grade. You can finally choose. For example, the range of difficulty (d ) for first grade is 0≤ d <0.35, the range of difficulty (d ) for second grade is 0.35≤ d <0.45, and the range of difficulty (d) for third grade is 0.45≤ d <0.455, the range of difficulty (d ) for 4th grade is 0.455≤ d <0.53, the range of difficulty (d ) for 5th grade is 0.53≤ d <0.57, and the range of difficulty (d) for 6th grade is 0.57 ≤ d <0.65, but is not limited thereto.

10 is a chart showing the number of learning target kanji selected for each grade and a preset difficulty range for each grade for classifying kanji for learning by grade.

Referring to FIG. 10, according to an embodiment of the present application, the Chinese character difficulty measuring device 100 selects 35 Chinese characters for first grade learning among predetermined Chinese characters 2, and 32 Chinese characters for second grade learning. You can select 108 Chinese characters for 3rd grade learning, 122 Chinese characters for 4th grade learning, 148 Chinese characters for 5th grade learning, and 127 Chinese characters for 6th grade learning.

11 is an experimental example linked to a device for measuring Chinese character difficulty using a regression analysis according to an embodiment of the present application, and is a chart showing the result of classifying Chinese characters to be studied for a first grader of elementary school.

Referring to FIG. 11, it is confirmed that the target Chinese characters that meet the general learning level in the corresponding grade were selected through the classification of the difficulty level in consideration of the frequency of use of Chinese characters, the frequency of Korean meaning, and the number of strokes of the Chinese characters in the sample text data (1). I can.

12 is a schematic configuration diagram of a Chinese character difficulty measuring apparatus using regression analysis according to an embodiment of the present application.

Referring to FIG. 12, the Chinese character difficulty measuring apparatus 100 includes a Chinese character utilization frequency obtaining unit 110, a Korean meaning frequency obtaining unit 120, a difficulty level analyzing unit 130, and a learning target Chinese character determining unit 140. can do.

The Chinese character utilization frequency acquisition unit 110 may prepare a first data set A for the Chinese character utilization frequency from the sample text data 1 for each of the predetermined Chinese characters 2.

Here, the sample text data 1 may include elementary textbooks for each grade, and the Chinese character utilization frequency acquisition unit 110 may prepare a first data set A divided for each grade.

The Hangul semantic frequency acquisition unit 120 may prepare a second data set B for the Hangul semantic frequency from the sample text data 1 for each of the predetermined Chinese characters 2.

Here, the sample text data 1 may include elementary textbooks for each grade, and the Hangul semantic frequency acquisition unit 120 may prepare a second data set B divided for each grade.

The difficulty level analysis unit 130 performs multiple regression analysis based on the first data set (A), the second data set (B), and the stroke number information (C) of each of the predetermined Chinese characters (2). ) Each difficulty level can be measured.

In addition, when the difficulty level analysis unit 130 receives appropriate learning timing information for each of the predetermined Chinese characters 2, the difficulty level may be measured by further considering the appropriate learning timing information.

The learning target Chinese character determination unit 140 may classify some of the predetermined Chinese characters 2 as learning target Chinese characters corresponding to a predetermined grade based on the difficulty level of each of the predetermined Chinese characters 2 and the frequency of use of the Chinese characters.

Hereinafter, based on the details described above, the operation flow of the present application will be briefly described.

13 is a flowchart illustrating a method of measuring Chinese character difficulty using regression analysis according to an exemplary embodiment of the present application.

The method of measuring Chinese character difficulty using regression analysis according to an embodiment of the present application illustrated in FIG. 13 may be performed by the apparatus 100 for measuring Chinese character difficulty described above. Accordingly, even if omitted below, the description of the Chinese character difficulty measuring apparatus 100 may be equally applied to the description of a method of measuring Chinese character difficulty using regression analysis according to an exemplary embodiment of the present disclosure.

Referring to FIG. 13, in step S1310, the Chinese character utilization frequency acquisition unit 110 prepares a first data set (A) for the Chinese character utilization frequency from the sample text data (1) for each of the predetermined Chinese characters (2). can do.

Specifically, in step S1310, the Chinese character utilization frequency acquisition unit 110 may collect sentences appearing in the sample text data 1. Subsequently, the Chinese character utilization frequency acquisition unit 110 writes a Chinese character corresponding to the Hangul part corresponding to the Chinese character word in each of the collected sentences, and maps each sentence with Chinese character notation and a predetermined Chinese character (2) to utilize Chinese characters. The first data set A can be prepared by counting the frequency.

Next, in step S1320, the Hangul semantic frequency acquisition unit 120 may prepare a second data set (B) for the Hangul semantic frequency from the sample text data 1 for each of the predetermined Chinese characters (2). .

Specifically, in step S1320, the Hangul semantic frequency acquisition unit 120 may tokenize the collected sentences and stemming the segmented words through tokenization. In addition, the Hangul semantic frequency acquisition unit 120 may prepare a second data set B by mapping a root converted through rooting and a predetermined Chinese character 2 to count the Hangul semantic frequency. Here, the Hangul semantic frequency acquisition unit 120 may count the Hangul semantic frequency based on the main word included in the converted root. In addition, the Hangul semantic frequency acquisition unit 120 may perform mapping in consideration of the synonyms of each of the predetermined Chinese characters 2.

Next, in step S1330, the difficulty analysis unit 130 performs multiple regression analysis based on the stroke number information (C) of each of the first data set (A), the second data set (B), and a predetermined Chinese character (2). Through this, the difficulty level of each of the predetermined Chinese characters (2) can be measured.

Specifically, in step S1330, when receiving appropriate learning timing information for each of the predetermined Chinese characters 2, the difficulty level analysis unit 130 may further consider the received appropriate learning timing information to measure the difficulty.

Next, in step S1340, the difficulty level analysis unit 130 may measure the difficulty level of a word made of Chinese characters based on the difficulty level of each of the predetermined Chinese characters. According to an exemplary embodiment of the present disclosure, in step S1340, the difficulty analysis unit 130 may determine an average value of the measured difficulty for each Chinese character constituting the word as the difficulty of the word.

Next, in step S1350, the learning target kanji determination unit 140, based on the difficulty level of each of the kanji (2) and the kanji utilization frequency, converts some of the kanji (2) to be studied kanji corresponding to a predetermined grade. It can be classified as

Specifically, in step S1350, the learning target kanji determination unit 140 filters kanji whose utilization frequency of kanji in all grades is greater than or equal to a preset threshold frequency among predetermined kanji (2) based on the first data set (A). I can. In addition, the learning target kanji determination unit 140, based on the first data set (A) prepared for each grade among the filtered kanji, the frequency of using kanji in a corresponding grade is greater than or equal to a preset threshold frequency for each grade, Chinese characters with difficulty levels that fall within the set difficulty range can be selected as the target Chinese characters for the grade.

In the above description, steps S1310 to S1350 may be further divided into additional steps or may be combined into fewer steps, according to an embodiment of the present disclosure. In addition, some steps may be omitted as necessary, or the order between steps may be changed.

The method for measuring Chinese character difficulty using regression analysis according to an exemplary embodiment of the present application may be implemented in the form of program instructions that can be executed through various computer means and recorded in a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, and the like alone or in combination. The program instructions recorded on the medium may be specially designed and configured for the present invention, or may be known and usable to those skilled in computer software. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tapes, optical media such as CD-ROMs and DVDs, and magnetic media such as floptical disks. -A hardware device specially configured to store and execute program instructions such as magneto-optical media, and ROM, RAM, flash memory, and the like. Examples of program instructions include not only machine language codes such as those produced by a compiler, but also high-level language codes that can be executed by a computer using an interpreter or the like. The hardware device described above may be configured to operate as one or more software modules to perform the operation of the present invention, and vice versa.

In addition, the method of measuring Chinese character difficulty using the regression analysis described above may be implemented in the form of a computer program or application executed by a computer stored in a recording medium.

The foregoing description of the present application is for illustrative purposes only, and those of ordinary skill in the art to which the present application pertains will be able to understand that it can be easily modified into other specific forms without changing the technical spirit or essential features of the present application. Therefore, it should be understood that the embodiments described above are illustrative and non-limiting in all respects. For example, each component described as a single type may be implemented in a distributed manner, and similarly, components described as being distributed may also be implemented in a combined form.

The scope of the present application is indicated by the claims to be described later rather than the detailed description, and all changes or modified forms derived from the meaning and scope of the claims and their equivalent concepts should be interpreted as being included in the scope of the present application.

10: Chinese character difficulty measuring system
100: Chinese character difficulty measuring device using regression analysis
110: Chinese character utilization frequency acquisition unit
120: Hangul meaning frequency acquisition unit
130: difficulty analysis unit
140: Chinese character determination unit to be studied
1: sample text data
2: prescribed kanji
A: first data set
B: second data set
C: Stroke information

Claims (15)

In the method of measuring the difficulty of Chinese characters using regression analysis,
For each predetermined Chinese character, preparing a first data set for a frequency of using Chinese characters from sample text data;
For each of the predetermined Chinese characters, preparing a second data set for the Hangul meaning frequency from the sample text data; And
Measuring a difficulty level of each of the predetermined Chinese characters through multiple regression analysis based on the first data set, the second data set, and the stroke number information of each of the predetermined Chinese characters,
Containing, Chinese character difficulty measuring method. The method of claim 1,
Preparing the first data set,
Collecting sentences appearing in the sample text data;
Stating a Chinese character notation corresponding to a Hangul part corresponding to a Chinese character word in each of the sentences; And
Counting the frequency of use of the Chinese characters by mapping each sentence with the Chinese character notation and each of the predetermined Chinese characters,
That includes, Chinese character difficulty measuring method. The method of claim 2,
Preparing the second data set,
Tokenizing the collected sentences;
Rooting the segmented words through the tokenization; And
Counting the Hangul semantic frequency by mapping each of the roots converted through the rooting and the predetermined Chinese characters,
That includes, Chinese character difficulty measuring method. The method of claim 3,
Counting the Hangul meaning frequency,
To count the frequency of the Hangul meaning based on the main word included in the converted root, Chinese character difficulty measuring method. The method of claim 4,
Counting the Hangul meaning frequency,
To perform the mapping in consideration of the synonyms of each of the predetermined Chinese characters, the Chinese character difficulty measuring method. The method of claim 3,
The sample text data includes elementary textbooks by grade,
The first data set and the second data set, characterized in that the preparation is divided by the grade, the Chinese character difficulty measuring method. The method of claim 6,
Measuring the difficulty of each of the predetermined Chinese characters,
When receiving information about an appropriate learning time for each of the predetermined Chinese characters, the difficulty level is measured by further considering the appropriate learning time information. The method of claim 6,
The method further comprising the step of measuring a difficulty level of a word made of Chinese characters based on the difficulty level of each of the predetermined Chinese characters. The method of claim 8,
Measuring the difficulty of the word,
To determine the average value of the difficulty measured for each Chinese character constituting the word as the difficulty of the word, Chinese character difficulty measuring method. The method of claim 6,
The method further comprising classifying some of the predetermined Chinese characters into learning target Chinese characters corresponding to a predetermined grade based on the difficulty level of each of the predetermined Chinese characters and the frequency of using the Chinese characters. The method of claim 10,
The step of classifying the learning target Chinese characters,
Filtering a Chinese character whose utilization frequency of the Chinese character in all grades is greater than or equal to a preset threshold frequency among the predetermined Chinese characters based on the first data set; And
Among the filtered Chinese characters, based on the first data set prepared for each grade, the frequency of using the Chinese characters in the corresponding grade is greater than or equal to a preset threshold frequency for each grade, and corresponding to a Chinese character having a difficulty level that falls within a preset difficulty range for the corresponding grade. Selecting the Chinese character for the study target for the grade,
That includes, Chinese character difficulty measuring method. In the Chinese character difficulty measuring device using regression analysis,
For each predetermined Chinese character, a Chinese character utilization frequency obtaining unit for preparing a first data set for the Chinese character utilization frequency from sample text data;
A Hangul semantic frequency acquisition unit for preparing a second data set for Hangul semantic frequencies from the sample text data for each of the predetermined Chinese characters; And
Difficulty analysis unit for measuring the difficulty of each of the predetermined Chinese characters through multiple regression analysis based on the first data set, the second data set, and the stroke number information of each of the predetermined Chinese characters,
Containing, Chinese character difficulty measuring device. The method of claim 12,
The sample text data includes elementary textbooks by grade,
The Chinese character utilization frequency acquisition unit prepares the first data set divided by grade,
The Hangul semantic frequency acquisition unit, characterized in that to prepare the second data set divided by grade, Chinese character difficulty measuring apparatus. The method of claim 12,
The difficulty analysis unit,
When receiving information about the appropriate learning time for each of the predetermined Chinese characters, the difficulty level is measured by further considering the appropriate learning time information. The method of claim 13,
The Chinese character difficulty measuring apparatus further comprises a learning target Chinese character determination unit for classifying some of the predetermined Chinese characters into learning target Chinese characters corresponding to a predetermined grade based on the difficulty of each of the predetermined Chinese characters and the frequency of using the Chinese characters.

Images