JP3733425B2 - Language acquisition support device, language acquisition support program, and language acquisition support method - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a language acquisition support apparatus that is useful for learning a language by applying ciliary play, and a program and method therefor.
[0002]
[Prior art]
Ciritri play, which everyone has become accustomed to since childhood, exists in many countries and has the aspect of simply enjoying the fun of the game.Children can learn their native language words and foreigners can learn other languages. It has an educational aspect that is useful for learning, and an academic aspect that is a subject of research on the theme of communication between humans and machines. In recent years, for example, serial game software has been developed for the purpose of assisting language acquisition, which is an educational aspect. In addition, research on when humans feel fun can also be found that even if the opponent of the siritori is siritori game software, it feels interesting when it is taught that it is a human being. ing.
[0003]
[Problems to be solved by the invention]
By the way, there are no words that have the word “n” at the beginning of the word in Japanese. There are certain rules for each language. Under such rules, there are advantages and disadvantages that depend on characters, such that there are many words that start with one letter but few words that start with another letter. The game is advanced so that it is advantageous to the self. However, since it is not easy to make the advantages and disadvantages of each character into quantitative statistical data, even in the development and research of the above-mentioned siritri game software, simply collecting a large number of vocabulary, It just uses the result of extracting the ending character.
[0004]
Therefore, the present invention provides an apparatus, a program, and a method that are useful for language acquisition using a citrus based on a quantitative research study on which strategy will be advantageous or disadvantageous in the citrine. To do.
[0005]
[Means for Solving the Problems]
  The present invention is a language based on a rule based on a predetermined citrine that outputs an answer word other than an already input or output word, which is different from the presented word with the ending character of the input presented word as an initial character. Acquisition support device A2An input receiving unit 1 that receives an input of a presentation word, and a candidate that excludes a word that has already been input or has already been output based on the ending character of the presentation word that has received the input. Candidate word extraction unit 2 for automatically extracting words from a predetermined word dictionary DB1 and extraction by candidate word extraction unit 2 based on a plurality of quantitative statistical data DB2 for winning the ritual for each character included in the word dictionary DB1 A preference level calculation unit 3 that calculates a preference level indicating the degree of advantage in a character-dependent ritual for each candidate word, an answer word determination unit 4 that determines an answer word, and answers the determined candidate word An answer word output unit 5 for outputting as wordsIn addition to the configuration of the language acquisition support apparatus A1 (FIG. 1), the frequency such as the number of sentences in which the presented word appears in the predetermined corpus DB 3 storing the natural language sentence, and the presented word and each candidate word in the corpus DB 2 An association level calculation unit 6 that calculates an association level obtained by a predetermined calculation formula including the frequency of occurrence of both of them is further included. In the answer word determination unit 4, preference levels and associations that are calculation results of the preference level calculation unit 3 The degree of goodness for each candidate word by an arithmetic expression in which the larger the value of the association degree that is the calculation result of the degree calculation unit 6 is, the larger value is obtainedThe relevantGoodThe candidate word with the highest degree is determined as the answer wordConfigure toIs a basic feature(Figure 2). Such language acquisition support device A2Realizes the intended function by operating a dedicated computer or a general-purpose computer in accordance with a program based on a predetermined method.
[0006]
  In other words, when the user and the computer, which is a language acquisition support device, proceed alternately with the citation, the highest preference calculated to win the citation with the last letter as the initial letter for the presented word entered by the user Every timeAnd the highest degree of association calculated from the viewpoint of human ease of association with wordsHaving the bestGoodnessThe language acquisition support device outputs the answer word. Therefore, a series of citrines is established by repeating the input of the presented word by the user and the corresponding language acquisition support device outputting the answer word, and the user is output in addition to the vocabulary possessed by the user. You will be able to learn while enjoying the word. It should be noted that the rules of citrine can be determined as appropriate before starting a series of citrines, and the word dictionary and the corresponding statistical data are suitable as long as they can be built in or connected to a computer. Things can be used. In addition, when calculating the preference level, the words that have already been input and the words that have already been output are excluded from the quantitative statistical data, or the calculation is performed without excluding those words. Variations are possible.
[0007]
  By the way, as long as a word that can be output as a reply word remains in the word dictionary, the language acquisition support device that is a computer adopts a strategy that is advantageous for the computer itself and disadvantageous for the other user. This makes it possible to advance the clitoris so as not to lose to a human user. Therefore, since the present invention uses quantitative statistical data based on a word dictionary, a word that can be used by a user by continuously outputting words having the same ending character as answer words using the quantitative statistical data. Specifically, the following approach is adopted as a basic tactical approach to reducingis doing. First, the quantitative statistical data DB2 has a candidate word ending character as the initial character with respect to the number of words having the candidate word ending character as the initial character, and has a character that means losing in the siritri rule as an initial character. As the preference element value having the ratio of the number of words, the preference degree obtained by an arithmetic expression indicating that the preference word has a greater degree of advantage for a candidate word having a larger preference element value. It is conceivable to perform calculation. In other words, as the number of words that end with a character that means losing is larger than the total number of words that start with a certain character, the other user answers the character that means losing as the ending. This is advantageous for the computer because it is easy to input a word to be entered. In addition, the character that means losing in the rules of Ciritri is, for example, the character “n” in principle in Japanese, but any rule that loses when ending with other characters. Can be set.
[0008]
  MoreInIn the present inventionQuantitative statistical data DB2 as having as preference element value the number of initial letter types of words in a set of words consisting only of two letters with the initial letter of the candidate word as the initial letter, A method of calculating a preference level obtained by an arithmetic expression indicating that a preference word has a higher degree of advantage in a preference word in the preference level calculation unit 3Is adopted. That is, since a relatively short word of two letters is very easy to recall for human beings, it becomes more advantageous to reply with a two-letter word having various letters at the beginning.
[0009]
  In the present invention, in addition to the above embodiments,The quantitative statistical data DB2 is assumed to have, as a preference element value, the number of types of initial letters of the words in a set of words having the ending letter of the candidate word as a ending letter. You may make it calculate the preference degree obtained by the computing equation which shows that the degree of advantage is so large that the candidate word with a larger value of a preference element value. With this method, it is possible to answer with words starting with various characters, so that it is easy to continue “attack” against the user.
[0010]
Further, the quantitative statistical data DB2 uses the ending character of the candidate word for the number of words having the ending character of the candidate word as the initial character and having no ending character as the ending character in the siritri rules. As a preference element value having a ratio of the number of words, the preference degree obtained by an arithmetic expression indicating that a preference word has a greater degree of advantage in a candidate word having a larger preference element value. A calculation method can be adopted. In this case, if there are not many types of characters at the end of the characters compared to the types of characters at the beginning of the word, the “attack” against the user will not continue.
[0011]
Furthermore, the quantitative statistical data DB2 has the number of words having the initial character of the candidate word as the end character as a preference element value, and the preference value calculation unit 3 has a small preference element value. If a strategy is used to calculate the preference obtained with an arithmetic expression that indicates that the degree of advantage is greater for a word, the answer word with the same letter as the end word of the presented word will be cut back and output. Would be hesitant to recall the next word presented.
[0012]
In addition, it is assumed that the quantitative statistical data DB2 has, as the preference element value, the number of words whose initial word is the suffix of the candidate word. In the preference calculation unit 3, the candidate word having a small preference element value You may make it calculate the preference degree obtained with the calculating formula which shows that the degree of advantage is so large. That is, when a word with few types of ending characters is output as an answer word, the user inevitably needs to consider the next word from among words having a few initial characters, which is “attack” for the user. It is because it becomes difficult to continue.
[0013]
  Also,The quantitative statistical data DB2 is further assumed to have as a preference element value the number of words having the ending of the candidate word as a prefix, and the preference calculation unit 3 has a large value of the preference element value. You may make it calculate the preference degree obtained with the calculating formula which shows that the degree of advantage is so large that it is a candidate word.
[0014]
Furthermore, there are cases where the field of words that can be used in a series of citrines is limited to a predetermined one. In order to achieve this appropriately, for example, the language acquisition support device A3 shown in FIG. In addition to the basic configuration of the language acquisition support device A1, by setting a plurality of fields of a word group that can be used as a rule of the citrine, and accepting an input of a field change command for designating any field It is preferable to further provide a word field changing unit 7 that changes the field of the word group in a series of series. In particular, when the function of the word field changing unit 7 is added to the language acquisition support apparatus A2 shown in FIG. 2, as shown in FIG. 3, the function of the word field changing unit 7 is changed to a field changing command that accepts input. Based on at least one of a plurality of word dictionaries set in advance for each field, a plurality of statistical data, or a plurality of corpora storing natural language sentences, a word dictionary, statistical data, or corpus corresponding to the corresponding word dictionary It may be selected.
[0015]
Also, depending on the user's language acquisition level, it may be desirable to be able to change the difficulty level in a series of series. In this case, as in the language acquisition support device A3 shown in FIG. 3, a plurality of difficulty levels that can be used as the rules of the citrine are set, and an input of a difficulty level change command that specifies one of the difficulty levels is accepted. Therefore, it is desirable to further provide a difficulty level changing unit 8 for changing the difficulty level in a series of series.
[0016]
As a particularly preferable function of the difficulty level changing unit 8, a word dictionary corresponding to a difficulty level change command received from a plurality of word dictionaries DB 1 and statistical data DB 2 respectively corresponding to a plurality of preset difficulty levels Examples include selecting DB1 and statistical data DB2. The difficulty level changing unit 8 may change the arithmetic expression for calculating the preference level according to the difficulty level changing command that has received the input. Further, when the function of the word field changing unit 7 is added to the language acquisition support apparatus A2 shown in FIG. 2, the difficulty level changing unit 8 has a plurality of word dictionaries DB1 respectively corresponding to a plurality of preset difficulty levels. The word dictionary DB1 and the statistical data DB2 or the corpus DB3 corresponding to the difficulty level change command received from the plurality of statistical data DB2 or the plurality of corpus DB3 storing the natural language sentences may be selected. Note that the word dictionary DB1, the statistical data DB2, and the corpus DB3 used here may be the same as or different from the word dictionary DB1, the statistical data DB2, and the corpus DB3 used in the field changing unit 7. In addition, the difficulty level changing unit 8 can change the arithmetic expression for determining the answer word based on the preference level and the association level.
[0017]
In addition to the above, when the user does not come up with a presentation word to be input next corresponding to the answer word, it is thought that if the computer outputs information that is a hint of the presentation word, the usability of the language acquisition support device is improved. It is done. To that end, as in the language acquisition support devices A3 and A4 shown in FIG. 3 or FIG. 4, in addition to the basic configuration aspect described above or a configuration aspect applied thereto, a hint for requesting a hint related to a presentation word to be input Hint request receiving unit 9a that receives a request command, and a hint from a word that starts with the last letter of the answer word output by the answer word output unit immediately before based on the hint request command received by this hint request receiving unit 9a It is preferable to provide a hint generation / output unit 9b for generating and outputting.
[0018]
Further, when the user wants to know the output answer word, the user-friendliness is improved if the meaning is also output. For that purpose, in addition to the basic configuration mode described above or a configuration mode to which it is applied as in the language acquisition support devices A3 and A4 shown in FIG. 3 or FIG. 4, at least the response word output by the response word output unit A word meaning request receiving unit 10a that receives an input word meaning request command, and extracts and outputs the meaning of the answer word from a predetermined word meaning dictionary DB 4 based on the word meaning request command received by the word meaning request receiving unit 10a. It is desirable to further provide a word meaning extraction / output unit 10b. The word meaning dictionary DB4, the word dictionary DB1, and the corpus DB3 are preferably associated with each other so as to store a common word.
[0019]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, an embodiment of the present invention will be described with reference to the drawings.
[0020]
This embodiment is for supporting the acquisition of the language of the user using the citrine, and here, an example in the language acquisition support apparatus A3 shown in FIG. 3 described above is shown. That is, the language acquisition support device A3 of FIG. 3 has both a basic configuration and a plurality of auxiliary configurations in the language acquisition support device according to the present invention. In the following, a case where a Japanese word is used as the rule of the citrine will be described.
[0021]
First, as shown in FIG. 5, the language acquisition support apparatus A3 includes a CPU 101, a memory 102, a storage device 103 such as a hard disk (HDD), an input device 104 such as a mouse and a keyboard, an output device 105 such as a display, and the like. In addition to the above, the general-purpose personal computer stores a software program to be described later, and has a database device 106 built in the personal computer or externally connected. In addition to the above configuration, the language acquisition support device A3 is provided with a communication interface connectable to the Internet, a LAN, etc., and operates by accepting access from a personal computer used by an external user. In the following, a mode in which a single personal computer is used as the language acquisition support device A3 will be described.
[0022]
Next, the functional configuration of the language acquisition support apparatus A3 will be described with reference to FIG. As described above, the language acquisition support device A3 operates in accordance with the stored language acquisition support program, so that the input reception unit 1, the candidate word extraction unit 2, the preference calculation unit 3, the response word determination unit 4, the response word Output section 5, association degree calculation section 6, field change section 7, difficulty change section 8, hint request reception section 9a, hint generation / output section 9b, word meaning request reception section 10a, word meaning extraction / output section 10b Demonstrate the function.
[0023]
Further, the word dictionary DB1, the statistical data DB2, the corpus DB3, and the word meaning dictionary DB4 are stored in the database device 106. These are, for example, “Japanese word in general”, “fruit name”, “ They are stored in association with each other after being classified according to fields such as “name of country”, “economic term”, etc., or by difficulty level. Specifically, for example, FIG. 6 shows only a part of the word dictionary DB1 based on the “classified vocabulary table (National Institute for Japanese Language, Hidehide Publishing, 1964)” as an example. It belongs to the field of “Japanese words in general”, and is divided into seven pieces of information separated by “,” (reading, headword, classification number, first lower number of classification number, second lower number of classification number , Whether the sample usage frequency is a predetermined number (for example, 7) or more words). For each word dictionary DB1, quantitative statistical data for 44 types of characters (Hiragana) excluding Japanese “n” is created and stored as a statistical data DB2. FIG. 7 shows an example of statistical data DB2 based on a certain word dictionary DB1. In the figure, preferences 1 to 6 indicate preference elements Px (x = 1 to 6), respectively, and preference element values related to each character are shown. Here, it is assumed that the elements necessary for establishing a strategy to be advantageous in the citrine are determined as follows, and the preference element values for the respective preference elements Px are obtained according to arithmetic expressions using the elements. Shall be.
[0024]
That is, as the element, first, the number of words in the word dictionary DB1 having the character as the ending character is expressed as “END”. Further, the number of words having the character as the initial character is represented as “BEGIN”. However, this “BEGIN” is obtained by excluding a word whose final character is “n”, which means losing, in a Japanese series. Next, the number of words whose initial character is “n” and whose final character is “n” is expressed as “NN”. In addition, the number of words whose initial character is the initial character and whose initial character is the final character is represented as “SAME”. Further, the number of types of characters that can be the initial character in a set of words having the character as a ending character is represented as “BKIND”. The number of types of characters that can be the initial character in a set of two characters having the character as the ending character is represented as “BKIND2”. Based on these elements, a preference element value related to each preference element Px is obtained as described below. In the method described below, the basic idea of the basic strategy of Ciritori is adopted to guide the language acquisition support device A3 to be advantageous by processing such as reducing the number of words that can be used by the user who is the partner of the Ciritri. Is.
[0025]
Specifically, the preference element P1 related to the preference 1 includes the ending character of the candidate word with respect to the number of words having the ending character of the candidate word as the initial character and not having the character that means losing in the siritri rule as the initial character. , And the preference element value for each character is the total number of such candidate words in the formula {(END) / (BEGIN (END >> BEGIN))} It is calculated | required by calculating based on. The preference element P2 related to the preference 2 indicates the number of words having the initial character as the initial character, and the preference element value for each character is obtained by calculating {BEGIN} for the candidate word. It is done. The preference element P3 related to the preference 3 has the ending character of the candidate word as the initial character and the character “n” as the initial character when compared with the total number of words having the initial character of the candidate word as the initial character. The preference element value for each character is obtained by calculating the candidate word based on the formula {(NN) / (BEGIN + NN)}. The preference element P4 related to the preference 4 indicates the number of words having the ending character of the candidate word as the initial character and the ending character, and the preference element value for each character is the {SAME for the candidate word. } Is calculated. The preference element P5 related to the preference 5 indicates the number of types of ending characters of the words in the set of words having the ending characters of the candidate words as the ending characters, and the preference element value for each character is a candidate. It is obtained by computing {BKIND} for the word. The preference element P6 related to the preference 6 indicates the number of types of initial letters of the words in a set of words composed only of two letters having the ending letter of the candidate word as the ending letter. Is obtained by calculating {BKIND2} for the candidate word. FIG. 7 shows characters arranged in order from the character having the highest preference element value in preference 1.
[0026]
The corpus DB3 is a database of Japanese examples using words included in the corresponding word dictionary DB1, but all the words included in the word dictionary DB1 are not necessarily included in the corpus DB3. Not necessarily. The example stored in the corpus DB 3 is automatically subjected to morphological analysis processing, and the number of times each word appears in all the examples of the corpus DB 3 is calculated. Information on whether or not is recorded in the “sample use frequency” column in the corresponding word dictionary DB1. In the word dictionary DB1, it is possible to set so that only the above-mentioned predetermined number of words are used in the series. In addition, it is preferable that the corpus DB 3 includes words that are relatively easy for the user to think of depending on the level of the user. For example, if the user is an elementary school student, the corpus DB 3 includes examples of elementary school textbooks and readers. It is good to use what you did.
[0027]
Furthermore, the word meaning dictionary DB4 stores the meanings of the words included in the word dictionary DB1 in association with the words. The word semantic dictionary DB4 may be provided in a plurality corresponding to each word dictionary DB1, or all the words contained in all the word dictionaries DB1 that may be used are stored in one database. May be.
[0028]
Hereinafter, the functions of the respective units will be described in a random order.
[0029]
The field changing unit 7 determines a word dictionary DB1 and a statistical data DB2 to be used based on a field changing command that designates one field input by the user, and in some cases also determines a corpus DB3 and a word meaning dictionary DB4. By doing so, the field of words that can be used as a rule of a series is changed in a series of series. In order to allow the user to easily input a field change command, for example, a screen that allows one field to be selected from a plurality of fields is displayed on the display or the like that is the output device 105 in advance. Good.
[0030]
The difficulty level changing unit 8 determines a word dictionary DB1 and statistical data DB2 or a corpus DB3 to be used based on a difficulty level change command that designates one difficulty level input by the user, or a preference level described later. The arithmetic unit 3, the association degree arithmetic unit 6, or the answer word determination unit 4 performs a process of changing the arithmetic expression used in the process of determining the answer word, and determines the difficulty level of the word used in the series of citrines as a citrine rule. change. In order to allow the user to easily input the difficulty level change command, for example, a screen is displayed in advance on the display as the output device 105 so that one difficulty level can be selected from a plurality of difficulty levels. Good.
[0031]
The input receiving unit 1 performs a process of receiving the presentation word W input by the user. In order to allow the user to easily input the presented word W, for example, a presented word input field or the like may be displayed on the screen of the output device 105 or the like in advance.
[0032]
The candidate word extraction unit 2 searches the word dictionary DB1 based on the presentation word W received by the input reception unit 1, and selects a candidate word Wi that can be a candidate for a reply word with the ending character of the presentation word W as the initial character. Extract all. At that time, a presentation word that has been previously input in a series of citations and an answer word that has been output first are excluded from the candidate words Wi to be extracted. In addition, when the ending character of the presented word is a long sound, the ending character is assumed to be “U”, and when the ending character is a cloudy sound or a semi-voiced sound, the utterance is removed from the clouded sound or the semi-voiced sound. The candidate word extraction unit 2 can execute processing based on the rules of the citrine, such as being a character.
[0033]
The preference calculation unit 3 refers to the statistical data DB 2 for each of all candidate words Wi (i = 1, 2,... N) extracted by the candidate word extraction unit 2, and each preference element of preferences 1-6. By calculating the preference element values Px_i (x = 1 to 6) relating to P1 to P6 based on the above-described rules, the preference Pi relating to each candidate word is calculated using the equation [Pi = Cp1xP1_i + Cp2xP2_i + Cp3xP3_i + Cp4xP4_i + Cp5xP5_i + Cp6xP6_i]. Here, in the formula, Cpx (x = 1 to 6) is a predetermined constant that can be appropriately determined. Here, the arithmetic expression for obtaining the preference Pi is a simple linear sum function like the above arithmetic expression, but more complicated terms of quadratic terms, logarithmic terms (log P1_i) and reciprocal terms are used. A function to which (1 / P1_i) or a combination thereof is added may be used. When the difficulty level is changed by weighting the predetermined preference element Px in the difficulty level changing unit 8 described above, the constant Cpx in the arithmetic expression may be changed.
[0034]
The association degree calculation unit 6 refers to the corpus DB 3 for each of all candidate words Wi (i = 1, 2,... N) extracted by the candidate word extraction unit 2 and determines the ease of association of each candidate word. The association degree Qi shown is calculated by an arithmetic expression [Qi = log {freq (Wi, W) / freq (W)}]. Here, freq (W) represents the number of sentences in which the presented word W appears in the corpus DB3, and freq (Wi, W) represents the number of sentences in which both the presented word W and each candidate word Wi appear in the corpus DB3. Represents. This arithmetic expression may also be a more complex function or a simpler function, similar to the arithmetic expression for obtaining the preference Pi described above.
[0035]
The answer word determination unit 4 calculates the candidate word Wi by an arithmetic expression [Ri = Cp x Pi + Cq x Qi] based on the preference Pi and the association Qi, which are the calculation results in the preference calculation unit 3 and the association level calculation unit 6. The degree of goodness Ri for each is calculated. However, Cp and Cq mean predetermined constants determined as appropriate. And the candidate word which gave the highest favorableness degree Ri among each candidate word Wi is determined as the reply word Wans. Here, the value obtained by adding a value obtained by multiplying each of the preference level Pi and the association degree Qi by a constant is defined as the goodness level Ri, but it is better by a more complicated calculation formula like the calculation formula for obtaining the preference Pi. The degree Ri may be calculated. In the above-described difficulty level changing unit 8, when the difficulty level is changed by weighting either the preference level Pi or the association level Qi, either the preference level Pi or the association level Qi in the above formula or The arithmetic expression may be changed so that both are raised to the power.
[0036]
The answer word output unit 5 outputs the answer word Wans determined by the answer word determiner 4 by displaying it on the display device 105 such as a display and presents it to the user. Therefore, the user inputs a new word having the initial character as the initial character for the answer word Wans.
[0037]
The hint request receiving unit 9a, when the user does not come up with a presentation word suitable for the next input, the hint input by causing the user to click a predetermined button displayed on the display or the like as the output device 105. A request command is received, the answer word Wans output by the answer word output unit 5 immediately before is extracted from the answer word determination unit 4, and the answer word Wans is input to the input receiving unit 1 instead of the presented word. In this case, based on the answer word “Wans” input to the input receiving unit 1, the candidate word extraction unit 2, the preference calculation unit 3, the association degree calculation unit 6, and the answer word determination unit 4 are input through a series of processing. Determine the best word for the presentation word. Then, the hint generation / output unit 9 b generates a hint including a phrase that can be conceived of the word based on the determined word, and outputs the generated hint to the output device 105. This process can be executed by the following process, for example. That is, if the user does not come up with a word that begins with “ru” when the output answer word “Wans” is a word that ends with “ru”, the answer word determination unit 4 performs “roulette” according to the above processing. When the word “h” is determined, the hint generation / output unit 9b generates and outputs a hint “Lule ??”.
[0038]
When the user wants to know the meaning of the answer word “Wans”, the word meaning request receiving unit 10a receives the word meaning request command input by causing the user to click a predetermined button displayed on the display or the like as the output device 105. The word meaning extraction / output unit 10b searches the word meaning dictionary DB 4 based on the input word meaning request command, extracts the meaning of the corresponding word, and outputs it to the output device 105.
[0039]
Next, an operation flow when the language acquisition support apparatus A3 is used will be described with reference to flowcharts shown in FIGS.
[0040]
As a premise, the user directly operates the language acquisition support apparatus A3 as described above, but substantially the same operation is performed when the user uses another computer connected to the language acquisition support apparatus A3. Thus, the language acquisition support device A3 can be used. Further, it is assumed that the language acquisition support device A3 has already been activated and is available, and a predetermined initial screen is displayed on the display as the output device 105.
[0041]
First, as shown in FIG. 8, when the field change command is received by the user in a state where the initial screen is displayed (START) (step S1), the language acquisition support device A3 performs a predetermined word according to the field change command. The dictionary DB1 and the statistical data DB2 are set to at least those corresponding to the relevant field (step S2), or the corpus DB3 and the word meaning dictionary DB4 are further set to correspond to the relevant field (not shown). Also, as shown in FIG. 9, when the user receives a difficulty level change command in a state where the initial screen is displayed (START) (step S11), the language acquisition support device A3 performs predetermined processing according to the difficulty level change command. The word dictionary DB1 and the statistical data DB2 are set to correspond to at least the difficulty level (step S12), or the corpus DB3 and the word meaning dictionary DB4 are further set to correspond to the difficulty level (not shown). , Formula 1, Formula 2 or Formula 3 is changed to correspond to the difficulty level (not shown).
[0042]
And as shown in FIG. 10, language acquisition assistance apparatus A3 receives the input of the presentation word W by a user from this state (START) (step S31). Here, in the case where the user cannot think of the input word W and cannot input it, the input of the input word W is made later, or the user inputs an impossible word such as a coined word appropriately formed (step S31). NO), since the citrine cannot be continued, it ends without accepting the presented word, resulting in a “losing” of the user. On the other hand, when the presentation word W is received (step S31; YES), first, it is determined whether or not the ending character of the presentation word W is “n” (step S32). The candidate word Wi is extracted by searching the word dictionary DB1 and using the ending character of the presented word W as the initial character (step S33). Here, if there is no candidate word Wi that can be extracted in the word dictionary DB1 (step S33; NO), the language acquisition support apparatus A3 ends because the language acquisition support apparatus A3 cannot continue the citation. "Lose". On the other hand, if the candidate word Wi can be extracted (step S33; YES), the degree of preference Pi is calculated for each extracted candidate word Wi according to the equation 1 with reference to the statistical data DB2 (step S34). Further, with respect to each extracted candidate word Wi, the association degree Qi is calculated according to Equation 2 with reference to the corpus DB 3 (step S35). Then, based on the calculated preference level Pi and association level Qi, the goodness level Ri for each candidate word Wi is calculated according to the equation 3 (step S36). Next, the goodness level Ri for all candidate words Wi is compared (step S37), the candidate word that has obtained the highest goodness level Ri is determined as the answer word Wans (step S38), and this answer word Wans is output. (Step S39). Here, it is determined whether or not the ending character of the answer word “Wans” is “n” (step S40), and if it is not “n”, the above process is repeated, whereby a series of the user and the language acquisition support device A3. This is the case. When the ending character of the input presentation word is “n” (step S32; YES), or when the ending character of the answer word is “n” (step S40; YES), the rules of the ritual Since “losing”, the series of series ends (END). However, as a rule of the series, when a character other than “n” is used as the ending character, the character is the ending character of the presented word or the answer word in step S32 and step S40. It can be set to determine whether or not there is.
[0043]
When the user requests a hint, as shown in FIG. 11, when the language acquisition support apparatus A3 receives a hint request command from the user (step S41), the answer acquisition word Wans determined and output immediately before is extracted (step S41). Step S42), through a process similar to Steps S31 to S37 based on the ending character of the answer word Wans (Step S43), determine a word that can be a presentation word similarly to Step S38 (Step S44). A hint is generated based on the word (step S45) and the hint is output (step S46).
[0044]
On the other hand, when the user requests the meaning of the answer word “Wans”, as shown in FIG. 12, the language acquisition support apparatus A3 receives the word meaning request command from the user (step S51), and determines and outputs the answer word “Wans”. The word meaning dictionary DB 4 is searched based on the above, the meaning of the corresponding word is extracted (step S52), and it is output (step S53).
[0045]
As described above, according to the language acquisition method based on the citrus using the language acquisition support device A3 of the present embodiment, the user can appropriately learn Japanese words while feeling fun through a kind of game called iritri. . In particular, the language acquisition support apparatus A3 side determines the most suitable word for answering the presented word input by the user, that is, the best word for the purpose of winning in the ritual game. It is mechanical and cool because it uses six types of advantageous preferences and has the feature of being able to output the optimal word as an answer word, while also preferentially selecting words that are easily associated with humans. In addition to selecting a word, it has a configuration that takes into account the fun of being able to output an association that is close to the selection made by humans. Moreover, since the field of the citrine and the difficulty level are changed and the meanings of hints and words are presented to the user, the user's feeling of use can be preferably improved.
[0046]
The present invention is not limited to the embodiment described above. For example, it is not necessary to use all of the six preferences. For example, any one of them can be selectively used, and whether or not to use association is also arbitrary. Also, the arithmetic expression used until the answer word is determined can be changed as appropriate. Furthermore, field change, difficulty level change, hint output, and word meaning output are associated with the basic configuration of the present invention, so whether or not to apply these is also arbitrary. Other specific configurations are not limited to the above embodiment, and various modifications can be made without departing from the spirit of the present invention.
[0047]
【The invention's effect】
  As described in detail above, the present invention is very useful for learning a certain language, particularly words, through the play of citrine. That is, when a user enters a word, the most appropriate word for an answer with the word's suffixAs an answer word with the highest goodness obtained from the preference calculated as an element for winning in the citrine and the association calculated as an element of human associationSince it can be output, and by repeating this alternately between the user and the computer, a series of citrines can be established until it becomes a foul in the citrine rules. As a result, the user can learn many words output from the language, and as a result, can learn many words in the language used in the cerritory.
[Brief description of the drawings]
FIG. 1 is a functional configuration diagram showing a basic aspect of a language acquisition support apparatus according to the present invention.
FIG. 2 is a functional configuration diagram showing an applied aspect of the language acquisition support apparatus according to the present invention.
FIG. 3 is a functional configuration diagram showing another applied aspect of the language acquisition support apparatus according to the present invention.
FIG. 4 is a functional configuration diagram showing another applied aspect of the language acquisition support apparatus according to the present invention.
FIG. 5 is a schematic device configuration diagram of a computer constituting a language acquisition support apparatus according to an embodiment of the present invention.
FIG. 6 is an explanatory diagram showing a part of an example of a language dictionary used in the embodiment.
FIG. 7 is an explanatory diagram showing an example of statistical data used in the embodiment.
FIG. 8 is a flowchart showing a part of the operation of the language acquisition support apparatus according to the embodiment;
FIG. 9 is a flowchart showing a part of the operation of the language acquisition support apparatus according to the embodiment;
FIG. 10 is a flowchart showing a part of the operation of the language acquisition support apparatus according to the embodiment;
FIG. 11 is a flowchart showing a part of the operation of the language acquisition support apparatus according to the embodiment;
FIG. 12 is a flowchart showing a part of the operation of the language acquisition support apparatus according to the embodiment;
[Explanation of symbols]
A1, A2, A3, A4 ... Language acquisition support device
DB1 ... Word dictionary
DB2 ... statistical data
DB3 ... Corpus
DB4 ... Word meaning dictionary
1 ... Input reception part
2 ... Candidate word extraction unit
3 ... Preference calculation unit
4 ... Answer word decision section
5 ... Answer word output section
6 ... Association degree calculation unit
7. Field change department
8 ... Difficulty level change part
9a ... Hint request acceptance department
9b ... Hint generation / output unit
10a ... Word meaning request reception part
10b ... Word meaning extraction / output unit

Claims (13)

Language acquisition support device based on a rule relating to a predetermined citrine that outputs an answer word excluding a word that has already been accepted or outputted, unlike the presentation word, with the ending character of the inputted presentation word as the initial character Because
An input receiving unit for receiving an input of the presented word;
Candidate word extraction unit for automatically extracting candidate words from a predetermined word dictionary based on the ending character of the presented word received by the input receiving unit, with the ending character as the initial character and excluding the word that has already been input or has already been output When,
Preference indicating the degree of advantage in the citation that depends on the character for each candidate word extracted by the candidate word extraction unit based on a plurality of quantitative statistical data for winning the citri for each character included in the word dictionary A preference calculation unit for calculating
An association degree calculation unit that calculates an association degree based on a value obtained by dividing a frequency at which a presentation word and each candidate word appear together in a predetermined corpus storing a natural language sentence by a frequency at which the presentation word appears in the corpus;
Each candidate word is calculated by an arithmetic expression that includes both the preference degree calculated by the preference degree calculation unit and the association degree obtained by the association degree calculation unit, and a larger value is obtained as the value of the preference degree and the association degree is larger. seek goodness, and answer word determiner that the goodness determines the maximum of the candidate words to answer word,
An answer word output unit that outputs the answer word determined by the answer word determination unit ,
The quantitative statistical data includes a word having a candidate word ending character as an initial character and a word having a character meaning losing in a siritri rule as the initial character with respect to the number of words having the candidate word ending character as an initial character. A preference ratio calculation unit having a ratio of numbers and the number of types of initial letters of those words in a set of candidate words composed of only two characters having the ending character of the candidate word as a ending character. However, the language acquisition support apparatus is characterized in that a preference level obtained by an arithmetic expression indicating that a candidate word having a higher preference element value has a higher degree of advantage is calculated . The quantitative statistical data further includes , as a preference element value, the number of types of initial letters of the words in a set of words having the suffix of the candidate word as a suffix , and a preference calculator calculates the preference 2. The language acquisition support apparatus according to claim 1, wherein a preference level obtained by an arithmetic expression indicating that a candidate word having a larger element value has a higher degree of advantage is calculated. The quantitative statistical data includes a word having a candidate word ending character as a ending character with respect to the number of words having a candidate word ending character as an initial character and not having a character meaning losing in the siritri rule as a ending character. The preference element value is further included as a preference element value, and the preference calculation unit calculates a preference degree obtained by an arithmetic expression indicating that a candidate word having a larger value of the preference element value has a higher degree of advantage. The language acquisition support apparatus according to claim 1, wherein the language acquisition support apparatus is one. The quantitative statistical data further includes , as a preference element value, the number of words having a candidate word ending character as an initial character and a ending character, and a preference calculation unit is a candidate whose value of the preference element value is small 4. The language acquisition support apparatus according to claim 1, wherein a preference level obtained by an arithmetic expression indicating that a word has a higher degree of advantage is calculated. The quantitative statistical data further includes , as a preference element value, the number of words having a candidate word ending as an initial character , and the preference calculation unit is more advantageous for a candidate word having a larger value of the preference element value. 5. The language acquisition support apparatus according to claim 1, wherein a preference degree obtained by an arithmetic expression indicating that the degree of the degree is large is calculated. A word field changing unit that sets the field of the word group that can be used as a rule of the ritual and changes the field of the word group in the series of rituals by receiving input of a field changing command that specifies any field The word field changing unit includes at least one of a plurality of word dictionaries, a plurality of statistical data, or a plurality of corpora that store natural language sentences set in advance for each field based on a field changing command that has received an input. 6. The language acquisition support apparatus according to claim 1 , wherein a word dictionary, statistical data, or corpus to be referred to is selected from any one of them . A plurality of difficulty levels that can be used as siritri rules are set, and a difficulty level change section for a series of serie is further provided by accepting an input of a difficulty level change command that designates one of the difficulty levels. A word corresponding to a difficulty level change command received from a plurality of word dictionaries and a plurality of statistical data corresponding to a plurality of preset difficulty levels, or a plurality of corpora storing natural language sentences. 7. A dictionary, statistical data, or a corpus is selected, or an arithmetic expression for determining an answer word based on a preference and association is modified. The language acquisition support device according to any one of the above. A hint request accepting unit that accepts a hint request command for requesting a hint related to a presentation word to be input, and a suffix of the answer word output by the answer word output unit immediately before based on the hint request command accepted by the hint request accepting unit The language acquisition support apparatus according to claim 1 , further comprising a hint generation / output unit that generates and outputs a hint from a word used as an initial character. . At least a word meaning request accepting unit that accepts an input word meaning request command regarding the answer word output by the answer word output unit, and a predetermined word meaning dictionary based on the word meaning request command accepted by the word meaning request accepting unit. The language acquisition support device according to claim 1 , further comprising a word meaning extraction / output unit that extracts and outputs the meaning of the answer word . By operating a computer, a predetermined citrine that outputs an answer word excluding a word that has already been accepted or output, unlike the presented word, with the initial letter of the presented word input to the computer as an initial character. A program for causing a computer to function as a language acquisition support device based on the rules related to
An input receiving means for receiving an input of the presented word;
Candidate word extracting means for automatically extracting candidate words from a predetermined word dictionary excluding words that have already been input or have already been input based on the ending character of the presented word received by the input receiving means. When,
Preference indicating the degree of advantage in the citation that depends on the character for each candidate word extracted by the candidate word extraction means, based on a plurality of quantitative statistical data for winning the citation for each character included in the word dictionary A preference calculation means for calculating
An associative degree calculating means for calculating an association degree based on a value obtained by dividing a frequency at which a presentation word and each candidate word appear together in a predetermined corpus storing a natural language sentence by a frequency at which the presentation word appears in the corpus;
Each candidate word is calculated by an arithmetic expression that includes both the preference level calculated by the preference level calculation unit and the association level obtained by the association level calculation unit, and a larger value is obtained as the value of the preference level and the association level is larger. An answer word determining means for obtaining a goodness degree and determining a candidate word having the maximum goodness degree as an answer word;
Function as an answer word output means for outputting the answer word determined by the answer word determination means,
The quantitative statistical data includes a word having a candidate word ending character as an initial character and a word having a character meaning losing in a siritri rule as the initial character with respect to the number of words having the candidate word ending character as an initial character. The preference element value is the ratio of the numbers and the number of types of initial letters of those words in a set of candidate words consisting only of two letters with the final letter of the candidate word as the final letter. A language acquisition unit, wherein the preference calculation means calculates a preference obtained by an arithmetic expression indicating that a candidate word having a larger value of the preference element value has a higher degree of advantage. Support program. The computer further accepts an input of a field change command that designates any field among a plurality of fields of the word group that can be used as a preset rule of the citrine, thereby determining the field of the word group in the series of citrines, It is changed by selecting a word dictionary, statistical data, or corpus to be referred to from at least one of a plurality of word dictionaries set in advance for each field, a plurality of statistical data, or a plurality of corpora storing natural language sentences. The language acquisition support program according to claim 10, which functions as a word field changing means. To a predetermined citrine that outputs a reply word excluding a word that has already been accepted or output, unlike the presented word, with the ending character of the presented word input to the computer as an initial character. A method for supporting language acquisition based on such rules,
An input receiving step for receiving an input of the presented word;
Candidate word extraction step for automatically extracting candidate words from a predetermined word dictionary excluding words that have already been input or have been output based on the ending character of the presented word received in this input reception step. When,
Preference indicating the degree of advantage in the citation that depends on the character for each candidate word extracted in the candidate word extraction step, based on a plurality of quantitative statistical data for winning the citation for each character included in the word dictionary A preference calculation step for calculating
An association degree calculating step for calculating an association degree based on a value obtained by dividing the frequency of occurrence of both the suggested word and each candidate word in a predetermined corpus storing a natural language sentence by the frequency of occurrence of the presented word in the corpus;
Each candidate word is calculated by an arithmetic expression that includes both the preference degree calculated in the preference degree calculation step and the association degree obtained in the association degree calculation step, and the larger the preference value and the association degree value, the larger the value is obtained. An answer word determination step for obtaining a goodness degree and determining a candidate word having the maximum goodness degree as an answer word;
An answer word output step for outputting the answer word determined in the answer word determination step,
The quantitative statistical data includes a word having a candidate word ending character as an initial character and a word having a character meaning losing in a siritri rule as the initial character with respect to the number of words having the candidate word ending character as an initial character. A preference ratio calculation means having a ratio of numbers and a number of types of initial letters of the words in a set of candidate words composed only of two characters having the ending characters of the candidate words as ending characters; However, the language acquisition support method is characterized in that the preference degree obtained by an arithmetic expression indicating that the candidate word having a larger preference element value has a higher degree of advantage is calculated. A plurality of fields of word groups that can be used as ritual rules are set in advance, and by receiving a field change command that designates one of the fields, a field of word groups in a series of citries is pre-defined for each field. A word field changing step that is changed by selecting a word dictionary, statistical data, or corpus to be referenced from at least one of a plurality of set word dictionaries, a plurality of statistical data, or a plurality of corpora storing natural language sentences. The language acquisition support method according to claim 12, further comprising:

Images