JP3700193B2 - Kana-kanji conversion device and kana-kanji conversion method - Google Patents

Description

[0001]
[Industrial application fields]
The present invention relates to a kana-kanji conversion device and a kana-kanji conversion method, and more particularly, to a kana-kanji conversion device and a kana-kanji conversion method used for selecting a kanji candidate for a word constituting a phrase using information on dependency between words. About.
[0002]
[Prior art]
2. Description of the Related Art Conventionally, various kana-kanji conversion devices that convert a kana character string input from a keyboard or the like into a desired kana-kanji mixed sentence have been proposed as a Japanese sentence input device or a Japanese sentence editing apparatus. In such a kana-kanji conversion device, it is not necessary for the user to specify each word or phrase separation position, and it is desired that the converted character string has the notation desired by the user. In Japan, there are many homophones and homonyms, so in order to obtain the correct kana-kanji mixed sentence without error, you will probably have to analyze the meaning of the sentence. In order to analyze the meaning, it is necessary to have a knowledge base of tens of thousands of words related at least organically, which is extremely difficult to realize.
[0003]
Therefore, the conventional Kana-Kanji conversion device attempts to obtain the desired result without analyzing the meaning by devising the phrase segmentation process and the learning process for selecting the homonym. As segmentation processing, a two-phrase longest match method in which two clauses are obtained as the first candidate among the phrases that can be composed of two clauses as a basic unit, or a word that can be a candidate for a word constituting the clause In addition, there is a minimum cost method in which a cost is given to a combination of words and a phrase whose score satisfies a predetermined condition is a first candidate. Also, in the learning process, learning of homophones with the highest priority given to the word selected by the user immediately prior to the homophone is used as the next candidate, or used as the length of the phrase containing a word. Learning the phrase length that gives the highest priority to the length specified by the person is known.
[0004]
Furthermore, recently, we have focused on specific relationships between words (for example, “hot” and “tea” in “hot tea” or “hot” and “summer” in “hot summer”) and remembered this relationship. By preparing a dictionary, when one word (for example, “tea”) is specified, a word related to this word (for example, “hot” among candidates for “hot”) is selected as the first candidate. Some have been proposed (for example, “Kana-Kanji conversion device” in Japanese Patent Laid-Open No. 3-105664, “Kana-Kanji conversion device” in Japanese Patent Laid-Open No. 4-277661, etc.). These specific relationships between words are called “dependency” or “co-occurrence”.
[0005]
[Problems to be solved by the invention]
However, the syntax rules of the languages used in reality are extremely complex, and simply referring to the dependency relationship does not give the desired Japanese language. What procedure does the dependency test do? It was a big problem. For example, when there are parentheses or punctuation marks in the input character string, there is no clear solution to the problem of how to test the dependency.
[0006]
In addition, when performing kana-kanji conversion for multiple phrases at once, where to check the dependency relationship, if multiple dependency relationships are found, what to do with the priority order, or even multiple dependency relationships Whether or not coexistence was allowed or not when the relationship was found was not made clear.
[0007]
The kana-kanji conversion apparatus and the kana-kanji conversion method of the present invention have been made for the purpose of clarifying these problems and obtaining a desired kana-kanji mixed sentence using the dependency relationship.
[0008]
[Means and Actions for Solving the Problems]
  The kana-kanji conversion device according to claim 1 is:
  A kana-kanji conversion device for inputting a kana character string and referring to a grammar dictionary to generate a kana-kanji mixed character string candidate,
  A segmentation means for segmenting the input kana character string with reference to the grammar dictionary;
  A dependency information dictionary storing dependency information between predetermined phrases;
  Dependency verification means for verifying the presence of a dependency with other clauses in order from the one having the smallest distance between phrases, starting from the one phrase that has been divided, and referring to the dependency information dictionary;
  Delimiter search means for searching for delimiters present in the input kana character string;
  A dependency test ending means for ending the dependency test starting from the one clause when the delimiter is found during the dependency test and using the position as the end position of the dependency test;
  WithIn addition,
The dependency verification means is:
Next candidate test means for sequentially determining the next candidate of the target phrase, with the test of the dependency test as a starting word for the phrase,
After the test by the next candidate test means is completed, the next candidate test means for continuing the test of the received word with the next candidate of the one phrase as the starting point as a related word;
  The main point is that
[0009]
  In this kana-kanji conversion device, the segmentation means refers to the grammar dictionary for the input kana character string, and segmentation is performed by referring to the grammar dictionary. With reference to a dependency information dictionary storing dependency information between clauses, the presence of dependency with other clauses is sequentially examined from the ones with the smallest distance between clauses from the starting clause. On the other hand, if the delimiter search means is searching for a delimiter that exists in the input kana character string and the delimiter is found during a dependency test starting from a certain phrase, The acceptance test end means uses the position as the end position of the dependency test and ends the dependency test starting from one phrase.In addition, the kana-kanji conversion device does not perform the dependency test only for the first candidate phrase, but also performs the next candidate, and gives priority to the test for the next candidate for the received word, Thereafter, the next candidate for the clerk is tested.
[0010]
  The kana-kanji conversion device according to claim 2 is:
  A kana-kanji conversion device for inputting a kana character string and referring to a grammar dictionary to generate a kana-kanji mixed character string candidate,
  A segmentation means for segmenting the input kana character string with reference to the grammar dictionary;
  A range search means for searching a range delimited by a pair of codes indicating start and end in the input kana character string;
  A dependency information dictionary storing dependency information between predetermined phrases;
  Dependency verification means for verifying the presence of a dependency with other clauses in order from the one having the smallest distance between phrases, starting from the one phrase that has been divided, and referring to the dependency information dictionary;
  A test range removing unit for removing the range searched by the range search unit from the range of the dependency test by the dependency test unit;
  WithIn addition,
The dependency verification means is:
Next candidate test means for sequentially determining the next candidate of the target phrase, with the test of the dependency test as a starting word for the phrase,
After the test by the next candidate test means is completed, the next candidate test means for continuing the test of the received word with the next candidate of the one phrase as the starting point as a related word;
  The main point is that
[0011]
  In this kana-kanji conversion device, the segmentation means refers to the grammar dictionary for the input kana character string, and segmentation is performed by referring to the grammar dictionary. With reference to a dependency information dictionary storing dependency information between clauses, the presence of dependency with other clauses is sequentially examined from the ones with the smallest distance between clauses from the starting clause. On the other hand, the range search means searches the input kana character string for a range delimited by a pair of signs indicating the start and end, and the test range removal means is determined by the dependency test means. Excluded from the scope of dependency testing. Therefore, the dependency test is performed by skipping a range surrounded by a pair of codes indicating start and end.In addition, the kana-kanji conversion device does not perform the dependency test only for the first candidate phrase, but also performs the next candidate, and gives priority to the test for the next candidate for the received word, Thereafter, the next candidate for the clerk is tested.
[0014]
  The kana-kanji conversion device according to claim 4 is:A kana-kanji conversion device according to claim 1 or 2, wherein
  When the segmented phrase is 3 or more, the one segment segmented as a starting point is referred to, and the presence of a dependency with another clause is referred to by referring to the dependency information dictionary. Dependency verification means to verify sequentially from the side near the end of the sentence
  The main point is that
[0015]
This kana-kanji conversion device determines the priority of dependency test when there are three or more shared phrases, and the dependency verification means starts with one shared phrase as a dependency information dictionary. The presence of a dependency with other clauses is sequentially examined from the side closer to the end of the sentence among three or more clauses.
[0016]
The kana-kanji conversion device according to claim 5 includes the dependency verification means according to claim 4,
A first test means for first performing a test starting from a related word close to the end of the sentence;
When a dependency is found by the first verification means, pay attention to the phrase ahead of the phrase that is the starting point, and only the dependency that is established with the dependency found by the first verification means. The second verification means to take out as the establishment of the reception
The main point is that
[0017]
Therefore, the kana-kanji conversion device first performs the dependency test near the end of the sentence first, and then only the dependency that has been established together with the dependency already verified in the preceding clause. We recognize as the establishment of
[0018]
  The kana-kanji conversion method of claim 6 is:
  A kana-kanji conversion method for inputting a kana character string and referring to a grammar dictionary to generate a kana-kanji mixed character string candidate,
  The input kana character string is segmented with reference to the grammar dictionary,
  Starting from one of the above-mentioned clauses, refer to the dependency information dictionary storing the dependency information between the predetermined clauses, and determine the presence of dependency with other clauses with a small distance between clauses And test sequentially from
  Search for a delimiter that exists in the input kana character string,
  When the delimiter is found during the dependency test, the dependency test starting from the one clause is terminated with the position as the end position of the dependency test.Shall be
Furthermore, in the dependency test,
One phrase as the starting point is used as a dependency word, and the dependency test of the dependency test is sequentially performed on the next candidate of the target phrase.
After the test for the next candidate is completed, continue the test of the spoken word using the next candidate of the one phrase as the starting point as a related word.about
  Is the gist.
[0019]
  The kana-kanji conversion method of claim 7 is:
  A kana-kanji conversion method for inputting a kana character string and referring to a grammar dictionary to generate a kana-kanji mixed character string candidate,
  In the input kana character string, a range delimited by a pair of codes indicating start and end is searched, and
  The input kana character string is segmented with reference to the grammar dictionary,
  With reference to the dependency information dictionary storing the dependency information of predetermined phrases while excluding the searched range from the dependency test range, starting from the one phrase that has been divided, and other Sequentially check the existence of the dependency on the clauses in ascending order of the distance between clauses.Shall be
Furthermore, in the dependency test,
One phrase as the starting point is used as a dependency word, and the dependency test of the dependency test is sequentially performed on the next candidate of the target phrase.
After the test for the next candidate is completed, continue the test of the spoken word using the next candidate of the one phrase as the starting point as a related word.about
  Is the gist.
[0020]
  The kana-kanji conversion method of claim 8 is:A kana-kanji conversion method according to claim 6 or 7,
  A kana-kanji conversion method for inputting a kana character string and referring to a grammar dictionary to generate a kana-kanji mixed character string candidate,
  The input kana character string is segmented with reference to the grammar dictionary,
  When there are three or more segmented phrases, the dependency information dictionary that stores the dependency information between predetermined phrases is used as a starting point for the one segment that has been segmented. Sequentially check for existence from the side closer to the end of the sentence among the three or more phrases
  Is the gist.
[0021]
【Example】
In order to further clarify the configuration and operation of the present invention described above, preferred embodiments of the present invention will be described below. FIG. 1 is a block diagram showing a control logic for kana-kanji conversion, and FIG. 2 is a block diagram showing hardware on which the kana-kanji conversion control logic actually operates. As shown in FIG. 2, this apparatus includes the following units connected to each other by a bus 31 around a known CPU 21. The units connected to each other by the CPU 21 and the bus 31 will be briefly described.
[0022]
ROM 22: mask memory for storing a kana-kanji conversion program, etc.
RAM 23: Readable and writable memory constituting main memory,
Keyboard interface 25: An interface for managing key inputs from the keyboard 24.
CRTC 27: CRT controller that controls signal output to the CRT 26 that can display in color.
Printer interface 29: an interface for controlling output of data to the printer 28;
A hard disk controller (HDC) 30; an interface for controlling the hard disk 32;
It is. The hard disk 32 stores various programs loaded into the RAM 23 and executed, a kana-kanji conversion processing program provided in the form of a device driver, or various conversion dictionaries referenced by the kana-kanji conversion processing program.
[0023]
With the hardware configured in this manner, text can be input, kana-kanji conversion, editing, display, printing, and the like. That is, the character string input from the keyboard 24 is subjected to predetermined processing by the CPU 21, stored in a predetermined area of the RAM 23, and displayed on the screen of the CRT 26 via the CRTC 27.
[0024]
Next, functions executed by the hardware thus configured will be described with reference to FIG. The configuration and operation of each unit shown in FIG. 1 will be outlined. The processing performed here is executed by the central processing unit (CPU 21) based on data input from the keyboard 24. All processing is performed by the CPU 21. As for kana-kanji conversion, a predetermined interrupt process is activated when the keyboard 24 is operated, and the input key image is converted into a corresponding kana character string, and further converted into a kana-kanji mixed character string. Starts. Of course, in a computer capable of parallel processing, kana-kanji conversion may be performed by one application (input method), and the conversion result may be transferred to a required application. In this case, inputs from the keyboard 24 are collectively accepted by the input method.
[0025]
The key image from the keyboard 24 is received by the character input unit 40 and is converted into a corresponding kana character string here. In the case of romaji input, it is converted into a kana character string with reference to a predetermined conversion table. Each time one kana character is obtained, the character input unit 40 sends the kana character to the conversion control unit 42. The conversion control unit 42 plays a central role in kana-kanji conversion, controls various kana-kanji conversion described later, and sends the result to the converted character string output unit 44. The post-conversion character string output unit 44 actually sends a signal to the CRTC 27 and displays the post-conversion character string on the CRT 26.
[0026]
The conversion control unit 42 passes the received kana character to the character string input unit 50. The character string input unit 50 stores the kana character string in the character storage unit 52. Based on this character string, the independent word candidate creation unit 54 and the adjunct word candidate creation unit 64 create word data candidates. The independent word candidate creation unit 54 uses the independent word dictionary 58 stored in advance in the hard disk 32 and performs processing for extracting independent word candidates from the obtained kana character string under the management of the independent word analysis position management unit 56. Do. On the other hand, the adjunct word candidate creation unit 64 uses the adjunct word dictionary 68 and performs processing for extracting an adjunct word candidate from the obtained kana character string under the management of the adjunct analysis position management unit 66. The process of extracting independent word candidates and auxiliary word candidates while moving the analysis position will be described later.
[0027]
Here, the independent word dictionary 58 changes priorities such as homonyms and affixes by learning. This learning process is performed by the dependency learning unit 70, the independent word learning unit 72, the auxiliary word learning unit 74, the affix learning unit 76, and the character conversion learning unit 78. The dependency learning unit 70 gives priority to the combination selected by the user in the same word combination when the user selects a word other than the word corresponding to the dependency under the condition that the dependency is satisfied. It learns dependency relationships. The independent word learning unit 72 learns the last selected word as the highest priority candidate in the independent word group in which the homonyms exist. The auxiliary word learning unit 74 learns which word form is used to convert an auxiliary word such as “please”, for example, “please” or “please”. Further, the affix learning unit 76 learns conversion formats (for example, “Go”, “Go”, etc.) such as prefixes and suffixes. The character conversion learning unit 78 learns a character string when the input character string is confirmed as hiragana or katakana as it is, and outputs the hiragana or katakana determined in the subsequent conversion processing as a candidate. .
[0028]
The independent word candidate creation unit 54 and the attached word candidate creation unit 64 obtain the created word candidates, and the word data creation unit 80 creates data for each word candidate. That is, as a result of the connection verification unit 82 referring to the connection verification table 84 for connection between the obtained independent words and ancillary words, independent words and independent words, and connection between phrases consisting of “independent words + ancillary words”, The result of the cost calculation unit 86 performing the overall cost calculation is obtained and output as data for each word. This word data is temporarily stored in the word data storage unit 100, receives the adjustment output from the dependency candidate adjustment unit 90, and is used for the phrase segmentation process.
[0029]
The dependency candidate adjustment unit 90 receives word candidates from the independent word candidate creation unit 54 and the adjunct word candidate creation unit 64 via the word data creation unit 80, the word data storage unit 100, and the phrase segmentation unit 102. The test is performed. The dependency test is performed by referring to a dependency dictionary 98 prepared in advance on the hard disk 32. Since the dependency dictionary stores only one dependency information even if the dependency relationship is reversed in order to reduce the capacity, the transposed information adjustment unit 99 is accompanied by a grammatical analysis. Thus, the information in the dependency dictionary 98 is expanded to adjust the dependency candidates. For example, only dependency information of the dependency word “Hanaga” + reception word “beautiful” is stored in the dependency dictionary 98, and an attempt is made to test the dependency of the dependency word “beautiful” + reception language “flower”. It is.
[0030]
The range in which the dependency test is performed is managed by the dependency range management unit 96. In addition, there are several allowable conditions for the dependency relationship test, which are determined by the service / passive analysis unit 92, the particle allowable analysis unit 94, and the like. The first phrase segmentation candidate is determined from the phrase candidates adjusted by the above dependency test, and is stored in the phrase data storage unit 106. The stored candidates are output to the conversion control unit 42 by the conversion character string output unit 108. The conversion control unit 42 displays this character string as a candidate character string, and an undesired character string may be a candidate. Therefore, in response to an instruction from the user, processing such as display and selection of the next candidate is performed. Do. These instructions, selection results, and the like are input to the phrase data storage unit 106 and the learning units 70 to 78 described above, and are used for determining part of the phrase, rewriting priority by learning, and the like. Although not shown, when the character string is finalized by the user, all data temporarily stored in each unit is deleted and prepared for the next conversion.
[0031]
The outline from the input of the kana character to the output of the conversion word character string has been outlined above. Next, details of each process will be described. First, general phrase segmentation processing will be described, and then dependency processing, which is a main part of the present invention, will be described. FIG. 3 is a flowchart showing an outline of phrase segmentation processing by the minimum cost method. As shown in the figure, first, after performing initialization processing (step S200) such as erasing temporarily stored data and initializing the analysis position to the first digit, processing for obtaining the analysis position is performed (step S200). Step S210). The analysis position is a position where the kana character string input so far is advanced one by one in order. For example, if the kana character string “Kagaku wa hako hakubu” is input as shown in FIG. 4, the first analysis position is the position of the first digit “ku”. At this analysis position, a process for searching the independent word dictionary 58 and the attached word dictionary 68 stored in the hard disk 32 is performed (step S220).
[0032]
After searching the dictionary, a process for checking the combination of the obtained word with the previous word is performed (step S230). If only words that cannot be combined between words are obtained, the dictionary is further stored. Search for. For example, in the example shown in FIG. 4, “wa” of the counsel retrieved from the adjunct dictionary 68 for “ha” of “kohahakobu” cannot be combined with the immediately preceding case particle “wo”. Therefore, it is treated as invalid data by the connection verification by the word data creation unit 80 and the connection verification unit 82. In FIG. 4, a symbol “x” is attached to a word that is determined to be invalid by such a combination check. The connection between words is stored in advance in the connection verification table 84. This connection verification table 84 is a table that gives information on the possibility of connection between parts of speech of a word. It is given as a matrix of × 400. When the dictionary search and the combination check at one analysis position are completed, the analysis position is advanced in order and the process is further repeated.
[0033]
For words that are likely to be combined, cost calculation is performed next, and processing for obtaining the minimum total cost of the words is performed (step S240). This process is performed by the cost calculation unit 86. In the example shown in FIG. 4A, “car” is, for example, “ku” + “ru” + “ma”, “car” + “ma”, “car”. , And when applying words to them, it is assumed that it has the cost of independent words = 2 and attached words = 0, and if it is “suffer” (independent words) + “style” (independent words), The total cost of “flow” is calculated as 4. At this time, the cost of “between” is 4 because the minimum total cost is calculated, and “come” + “between” is not the cost 6 of “bitter” + “flow” + “between”. This is because the case cost 4 is adopted. Since “de” and “ha” are attached words, the cost of the word “car” = 2 with the lowest cost among the previous words becomes its own cost. In FIG. 4, the cost of each word is shown in the lower right.
[0034]
After the above cost calculation, the cost of each word is checked, and a process of invalidating an inappropriate cost is performed (step S250). Inappropriate cost is a combination of words that is costly compared to other word combinations. In other words, selecting a combination of words such as “ku” + “toru” will be more expensive than other words “coming” and “repeating” obtained up to that position. Judgment is made and it is excluded from the phrase candidates. In FIG. 4, words that are not adopted based on this concept of minimum cost are shown as “●” in the upper right of the word. In FIG. 4, “◯” indicates that the word remains as a word that may form a phrase candidate as a result of the above-described combination check and cost check.
[0035]
Next, a process for linking the word candidates thus given the cost is performed (step S260). That is, for words for which coupling is valid, the coupling relationship is related by setting a pointer. In the example of FIG. 4, “come”, “carry”, “car”, “to”, “de”, “ha”, “ha”, etc. have been calculated as the minimum total cost. It links to “to” and “car” to “de” and “to”. This combination check, cost calculation, and linking process are repeated while searching for all words at one analysis position is completed. Further, when the dictionary search at the analysis position is completed, the analysis position is further advanced by one, the establishment of a new word is examined, and the connection check and cost calculation are repeated in the same manner.
[0036]
When the analysis position reaches the position of the last input kana character and the analysis is completed for all words (step S265), a process of searching for the path with the minimum cost is performed on the assumption of the above processing (step S265). Step S270). This process is performed by the phrase segmentation unit 102, and is a process of searching for a combination of valid words that minimizes the sum of the costs assigned to the words. In the example of “Humps until coming”, as shown by the solid line J in FIG. 4B, the division cost of “by car” + “carrying a box” + “carrying” is a total cost of 18, so the minimum Selected as a cost. Although not the minimum cost, other phrase segmentation candidates are also searched. For example, as shown by a broken line B in FIG. 4B, the division is “by car” + “child” + “carry” (cost = 20). In this way, after the candidate for the division is created (step S280), the process for creating the candidate inside each phrase is performed (step S290). That is, for example, candidates such as “box” and “box” are prepared for “hako wa” within one phrase segment. These phrase candidates and word candidates are used when the user is instructed to change the way of segmenting or to display the next candidate.
[0037]
Next, the dependency test process executed after the phrase segmentation process is performed and the kanji candidates are created for each phrase will be described. FIG. 5 is an explanatory diagram showing an example sentence to be converted to kana-kanji, FIG. 6 is a flowchart showing an end position search processing routine for searching for the end position of dependency, and FIG. 7 shows a dependency test processing routine. It is a flowchart to show. Here, as an example sentence, as shown in FIG. ”And a sentence with parentheses (“ ”) inside. It is assumed that the phrase shown in FIG. 5B is extracted by the division processing using the minimum cost method. First, an end position search processing routine for searching for a dependency end position will be described with reference to FIG.
[0038]
The end position search processing routine shown in FIG. 6 is a process for determining the range of the dependency test. In other words, when a plurality of clauses are extracted from the input kana character string by the phrase segmentation process, it is a process of determining from which clause to which clause it is desirable to perform the dependency test. When this processing routine is started, the first clause position of the input and segmented clause is first set as the dependency analysis range start position (step S300), and this clause is a clause generated from word data. It is determined whether or not there is (step S310). Among phrases, there are phrases consisting of indefinite words, such as “if X = A”. Since a phrase that does not include an independent word retrieved from the independent word dictionary 58 cannot be enough for a dependency to be established, it is not necessary to test the dependency range.
[0039]
If it is determined that the phrase is generated from the word data, then a process of extracting the last word in the phrase is performed (step S320). The phrase is configured as an independent word + an appendix, but a punctuation mark or the like may be attached at the end. Therefore, the last word taken out is the left parenthesis—for example, “(,“, [”, etc., judgment (step S330), right parenthesis—for example,“, ”,]]”, etc. Judgment (step S340), a punctuation mark-for example, ".・. ,? ! ] Is determined (step S350). If it is determined that none of these is found, or if the clause of interest is not a clause generated from word data in step S310, the dependency analysis range is searched for all the clauses created by the phrase segmentation. It is determined whether or not it has been performed (step S360). If all the phrases are not determined, the process moves to the next phrase (step S370), and the above processing is repeated from step S320.
[0040]
On the other hand, whether it is determined that the last word in the phrase is a left parenthesis, a right parenthesis, or a punctuation (steps S330, 340, 350), or if all the phrases are determined (step S360), The current phrase is determined to be the end of the dependency analysis range (step S380), and the process returns to "END" to end the present routine.
[0041]
With the above processing, for example, in the example sentence of FIG. 5, when the search for the end position is started from the phrase “red”, the last phrase “flower” is not included in the last word “red” because the last word “red” is not included. Will move to "", and since there is no left parenthesis or the like in the last word here, it moves to the next phrase "Utsui". Here, since the last word is a right parenthesis “” ”, it is determined that this is the end of the dependency analysis range. That is, first, ““ red flowers are beautiful ”” is determined as the dependency test range. Similarly, a judgment was made on the subsequent passages: 'I opened the window and I said. ] Is determined as the next test range for dependency.
[0042]
When the dependency test range is thus determined, the dependency test processing routine shown in FIG. 7 is started. In FIG. 6, the dependency test range is determined for all the phrases generated by the phrase segmentation process. However, each time a dependency test range is determined, the test processing routine of FIG. 7 is started. There is no problem.
[0043]
When the dependency test processing routine of FIG. 7 is started, first, a process is performed in which the phrase immediately before the end of the range defined as the test range is set as a dependency word (step S400). In the embodiment, the dependency is searched with the priority given to the dependency word. Therefore, the phrase immediately before the end of the test range is set as the phrase of the dependency word so as to have at least one reception word. . At this time, a process for initializing each variable of the process is also performed (for example, n is set to a value of 1). Next, a process of setting the n-th phrase after the dependency word as a received word is performed (step S410), and a process of searching the dependency dictionary 98 is performed (step S420). The dependency dictionary is recorded in the format of “reception word root” + “dependency word”.
[0044]
An example of the dictionary is shown in FIG. FIG. 8A schematically shows the contents of the dictionary in the case where there is a dependency of “promotion + working”. In this dictionary, the reading “Kikiten” as the headword and the words “handy” and “motivation” are registered, and if the dependency is “flower + beautiful”, as shown in FIG. The words “Utsushi Hana” and the words “beautiful” and “flower” are registered. At the end, attached word information for attached word allowable analysis described later is attached. The actual dictionary is accompanied by an index for searching, information indicating the length of words, and the like.
[0045]
As a result of searching the dependency dictionary from the words set as the dependency words and the words set as the reception words, whether or not a headline (for example, “Utsushi Hana”) created from the dependency words and the reception words exists in the dictionary If no dependency is found in the dictionary, it is determined whether or not the end of the test range has been tested (step S440). If not tested to the end, the variable n is set to the value 1 (Step S450), and the process is repeated from step S410. If a dependency is found in the dependency dictionary 98 (step S430), it is determined whether a dependency has already been established for the received word (step S460). As shown in FIG. 9, when the dependency Q1 + R2 is established once, the next time the dependency is determined for the previous phrase P, the process is changed so that the dependency Q1 + R2 already found is prioritized. Because.
[0046]
If no dependency has already been established on the receiver side (ie, word Q in the dependency determination between word P and word Q), the process proceeds to step S470 and subsequent steps, and the dependency is established first with dependency word priority. Search for the word you want. The reason for searching for a dependency-satisfied word with priority on the dependency word is that, according to the determination in step S430, it is known that there is a word that can have a dependency, and for which word the dependency is satisfied, The order on the side of the dependency word is judged first. The state of this search is shown in FIG. When a plurality of words are found for the reading on the side of the dependency word, the first candidate X1 is fixed, and the received word Y is determined according to the learning order already arranged in the independent word dictionary 58. In order from the highest order, it is checked whether the dependency is established in the order of Y1 → Y2 → Y3 → Y4... (Search for reference A1 in FIG. 10). If no word satisfying the dependency is found by this search, the next dependency word X2 is selected and the test is performed in the same manner (search for reference A2).
[0047]
When a combination of a dependency word and a reception word satisfying the dependency read from the dependency dictionary 98 is found by such a search, an admissible word admission analysis is performed (step S480). This process will be described.
[0048]
The admissible analysis of particles determines whether or not the allowable relationship defined by the dependency type is satisfied, and has the following types for each dependency type.
[I] Continuous modification type
▲ 1 Noun + particle + particle in the case of a predicate
Case particles "ga" "from" "de" "to" "ni" "to" "more" "to" "no"
Particle particle "ha"
(2) In the case of idioms + idioms
(3) Optional particle in the case of noun + predicate (particle abbreviation type)
"GA" "HA" particle, adjunct particle
[II] Complex modification type
(4) Noun + particle + particle in case of noun
"of"
(5) Particles in the case of body + body (parallel)
"Ya" "To"
(6) In the case of predicate form + noun
(7) In case of conjunction + noun
[0049]
That is, it is assumed that the relationship between two words determined to be in a dependency relationship belongs to any one of the above-mentioned (1) to (7), and an adjunct word existing between both words in the dependency relationship (mostly Is a particle or a particle expression), the dependency dictionary 98 is set with an allowable particle for a word having a dependency relationship, and this is verified. For example, when the dependency between “promotion” and “handedness” is accompanied by the allowable setting of the particle (no ·), it belongs to the case (1) above (noun + particle + probe). “” And “ga” can exist between the two words (smooth, savvy → ○), but other case particles “kara” and “de” are unacceptable. Would be good →→).
[0050]
Regarding the relations (1) to (7), it is judged that the relations other than those listed are allowed. Examples of determinations that are permitted are listed below, but these may include cases where the dependency does not hold as an actual expression. However, dependency is a broad concept as an actual human language activity, and a too strict dependency arrangement is often not suitable for reality. Also, a too strict dependency arrangement will only lead to a mischievous increase in the dependency dictionary 98 and will also reduce the speed of dependency verification. Therefore, in this embodiment, with respect to the admissibility of attached words, the relations in which dependency occurs are divided from (1) to (7), and those that are clearly permitted or not permitted are permitted in the dependency dictionary. It is memorized as a word with a dependency relationship, and the others are allowed.
[0051]
[III] Permissible expressions-in the case of continuous modification
・ Noun + case particle expression + case particle expression in predicate
“By”, “As”, “For”, “In”, “By”, etc.
・ Noun + particle + verb particle
“Some”, “Even”, “Shi”, “But”, “M”, etc.
・ Noun + adverb + adjunct in adjunct
"Kiri" "About" "One by one" "Only", etc.
・ Noun + adverbial expression + adverbial expression in predicate
"So," "If", etc.
・ Phrase + particle + particle
"Noha" etc.
・ Connecting particles “So” “From” “From” “Te”, etc.
・ Conjunctive particle-like expressions
・ Expression that parallels preaching + preaching “ka” “shi” “ri” “simultaneously”, etc.
[0052]
[IV] Permissible expressions-complex modifications
・ Noun + particle-like expression + particle-like expression in noun
"In", "related", "based on", etc.
・ Phrase + particle-like expression + particle-like expression in noun
"For", "like", "with", "etc."
・ Expression that parallels body language + body language "ka".
[0053]
According to the above rules, the admissibility of an adjunct word between two words for which a dependency relationship is found is determined. In the case of “flower” and “beautiful” given as an example, the allowable case particle is “ga”, and in this case, the formation of dependency is recognized. Therefore, this is determined (step S480), and when the dependency is established, the word for which the establishment is recognized is determined as the first candidate of the dependency word and the received word in the independent word constituting the phrase. Is performed (step S490). In other words, the order of registration by learning homophones in the independent word dictionary 58 is changed. Further, the range from the received word to the related word thus found is registered as a dependency established range, and a process for managing this is performed (step S500), and it is determined whether dependency search has been performed for the entire range ( Proceed to step S510). In addition, when the establishment of the dependency is denied due to an attached word existing between the two words, the first candidate is not changed, and if there is other dependency information about the dependency word and the received word, the same applies. If the test is performed (labor saving in the figure), the process proceeds to the determination as to whether or not the entire range has been completed.
[0054]
If the dependency test has not been completed for the entire dependency test range, that is, the entire dependency analysis range determined by the processing shown in FIG. 6, the variable n is initialized to a value 1 (step S520). Further, the dependency word is moved to the previous phrase (step S530), and the above processing (steps S410 to S500) is repeated until the dependency test is completed for the entire range.
[0055]
If it is determined in step S460 that the dependency has already been established for the received language, it is determined whether the dependency using the received language for which the dependency has already been established is established (step S460). S540). That is, in the example shown in FIG. 9A, when a dependency is established between the word Q1 and the word R2, when the dependency between the word P and the word Q is determined, The word Q1 that is the word is fixed, and the establishment of the dependency is determined. Accordingly, in this case, even if there is a dependency P1 + Q2, it is not adopted. However, if a dependency P2 + Q1 having a received language Q1 is found, this is a dependency. Therefore, as shown in FIG. 9 (B), the dependency of P2 + Q1 + R2 is established. In the example shown in FIG. 4, “flower” + “beautiful” is found for “red”, which is the dependency test range, and “flower” “beautiful” is the first candidate. After learning, the dependency “red” + “nose” exists, and this dependency is not adopted. On the other hand, a dependency of “red” + “flower” is adopted.
[0056]
Therefore, in this case, the first candidate for the related word (word P2 in the example) is determined (step S550). Thereafter, similarly, the management of the dependency range (step S500) and the determination as to whether the dependency test is completed for the entire range (step S510) are performed. When the dependency test has been completed for all ranges determined as the dependency test range, the process returns to “END” and the present routine ends.
[0057]
According to the present embodiment described above, the following effects can be obtained. First, the dependency test range is clarified. That is, the dependency is not determined beyond the left and right parentheses and punctuation marks, and the dependency can be determined in a range close to the range where the dependency extends in the normal syntax. Therefore, the range of the dependency is not increased excessively and the processing does not take excessive time. In the example sentence shown in Fig. 5, the dependency test range is "Akai Hanaga Tsukui" and "I opened the window and I said. Therefore, even if a dependency such as “beautiful” + “window” exists in the dependency dictionary 98, it is not verified as the presence of the dependency.
[0058]
In the present embodiment, the dependency test is started from a position close to the end of the sentence within the range of the dependency test range, and the dependency word is given priority. This configuration is extremely effective in making the determination of word candidates by dependency more desirable. This is because, in Japanese, the predicate side at the end of a sentence often takes on the meaning of the sentence, and the action (generally the predicate at the end of the sentence is the same) and the subject (generally the description at the beginning of the sentence is replaced) It is thought that this is because the actions of the subject often change.
[0059]
Further, when it is determined that the dependency has been established once, the range from the received word to the dependency word is managed as the dependency establishment range, so that the dependency ranges do not intersect. In addition, it is not determined that two or more spoken words receive a single dependent word. In addition, since the establishment of the dependency is judged beyond the adjacent clause (when n ≧ 2), the dependency test can be correctly performed even if the modification by the adverb or the like enters between the dependency relationships. Therefore, when a plurality of dependencies are established, as shown in FIG. 11A, a combination in which independent dependencies are established separately or as shown in FIG. Or a combination in which another dependency is established so as to straddle one dependency, as shown in FIG. 11 (C).
[0060]
Next, a second embodiment of the present invention will be described. The kana-kanji conversion apparatus of the second embodiment executes a dependency test range determination processing routine shown in FIG. 12 instead of the end position search processing routine for determining the dependency range shown in FIG. 6 of the first embodiment. The others are the same as those of the first embodiment. As shown in FIG. 12, the dependency test range determination processing routine of the second embodiment first searches all the clauses that have been input and subjected to the phrase segmentation processing (step S600), and forms a pair of codes, for example, ““ ] And "" "," ("and") ", etc. are determined (step S610). If they are present, the analysis target is moved to the inside (step S620), and there is no longer a pair of codes. For example, the range delimited by the punctuation marks (.?!, Etc., but not including the punctuation marks ",,") is determined as the dependency test range (step S630). On the other hand, for the outside of the range enclosed by the paired codes, the range before and after the range is integrated and the range delimited by the punctuation is determined as the dependency test range (step S640).
[0061]
That is, according to this processing, in the case of the example sentence “He is“ Natsu is hot ”and I did it” shown in FIG. 13, the dependency is made using the first candidate word obtained by the phrase segmentation. The test range is indicated by “Summer is in the paired symbol“ ””. "Thick" and the outside "he is When, to me Let go. And become ( Indicates the break between phrases), and the presence of the dependency B1, B2, B3, "the summer is hot" is learned as the first candidate in the former, And told me. Is learned as the first candidate. Therefore, it is possible to correctly determine the existence of a dependency so as to be established with a large node surrounded by parentheses.
[0062]
In the two embodiments described above, the dependency test is performed in order from the shortest distance between phrases starting from one phrase, but as shown in FIG. First, the test is performed for all the distances between phrases (1) (“C” + “D”, “B” + “C”, “A” + “B”), and then the distance between phrases (2) (“B”). ”+“ D ”,“ A ”+“ C ”), and further distant items (“ A ”+“ D ”) may be tested in order. Even in this case, as shown in FIG. 14, it is desirable to test from the dependency between the following clauses.
[0063]
Although the embodiments of the present invention have been described above, the present invention is not limited to these embodiments. For example, instead of the minimum cost method, another phrase segmentation method such as a two-segment longest match method is used. , A configuration that can switch between dependency word priority and reception word priority in dependency verification, a configuration that limits the maximum inter-sentence distance (in the embodiment, the size of n) for dependency verification to a predetermined size, dependency dictionary Needless to say, the present invention can be implemented in various modes within the scope of the present invention, such as other configurations.
[0064]
【The invention's effect】
  As described above, in the kana-kanji conversion device of claim 1 and the kana-kanji conversion method of claim 6 of the present invention, the input kana character string is segmented by referring to the grammar dictionary and segmented. The distance between phrases from a phrase that starts from a single phrase and refers to the dependency information dictionary that stores the dependency information between the specified phrases, and the existence of the dependency as another phrase Test sequentially from the smallest. On the other hand, if you search for a delimiter that exists in the input kana character string, and a delimiter is found during a dependency test starting from a phrase
Then, with the position as the end position of the dependency test, the dependency test starting from one phrase is ended.In addition, with the single phrase as the starting point as a dependency word, the dependency test of the dependency test is sequentially performed for the next candidate of the target phrase, and after the verification for the next candidate is completed, the determination is made as the starting point. Continue to test the spoken word using the next candidate of one phrase as a related word.Therefore, it is possible to appropriately determine the range for performing the dependency test in accordance with the Japanese syntax structure.In addition, the kana-kanji conversion device does not perform the dependency test only for the first candidate phrase, but also performs the next candidate, and gives priority to the test for the next candidate for the received word, Thereafter, since the test is performed on the next candidate for the clerk, the test on other candidates can be performed together.
[0065]
  In the kana-kanji conversion device according to claim 2 and the kana-kanji conversion method according to claim 7, the input kana character string is segmented by referring to the grammar dictionary, and one segmented phrase is used as a starting point. With reference to a dependency information dictionary storing dependency information between predetermined clauses, the presence of dependency with other clauses is sequentially examined from the ones with the smallest distance between phrases from the starting phrase. On the other hand, in the input kana character string, a range delimited by a pair of codes indicating start and end is searched, and this range is excluded from the range of dependency test.In addition, with the single phrase as the starting point as a dependency word, the dependency test of the dependency test is sequentially performed for the next candidate of the target phrase, and after the verification for the next candidate is completed, the determination is made as the starting point. Continue to test the spoken word using the next candidate of one phrase as a related word.Therefore, the dependency test is performed by skipping a range surrounded by a pair of codes indicating start and end. As a result, it is possible to test a dependency that is established beyond an insertion phrase or insertion clause.In addition, the kana-kanji conversion device does not perform the dependency test only for the first candidate phrase, but also performs the next candidate, and gives priority to the test for the next candidate for the received word, Thereafter, since the test is performed on the next candidate for the clerk, the test on other candidates can be performed together.
[0067]
In the kana-kanji conversion device of claim 4 and the kana-kanji conversion method of claim 8, the priority order of the dependency test is determined when there are three or more segmented phrases, and one segmented phrase is used as a starting point. Since the dependency information with other clauses is checked with reference to the reception information dictionary, the kana kanji with natural Japanese syntax using dependency is sequentially checked from the side closer to the end of the sentence. Candidates for conversion can be obtained.
[0068]
In the kana-kanji conversion device according to claim 5, first, the dependency test near the end of the sentence is performed first, and then only the dependency that is established together with the dependency whose dependency has already been verified in the preceding clause. Since the acceptance is accepted, it is possible to obtain candidates for kana-kanji conversion using natural Japanese syntax without forcing the dependency.
[Brief description of the drawings]
FIG. 1 is a functional block diagram showing an implementation form of a kana-kanji conversion function in a kana-kanji conversion apparatus according to an embodiment of the present invention.
FIG. 2 is a block diagram illustrating hardware that realizes a kana-kanji conversion device as an embodiment;
FIG. 3 is a flowchart showing a phrase segmentation process executed in the segment segmentation unit 102;
FIG. 4 is an explanatory diagram showing an example of phrase segmentation by a minimum cost method in the embodiment.
FIG. 5 is an explanatory diagram showing an example sentence that is a subject of a dependency test in the embodiment.
FIG. 6 is a flowchart showing an end position search processing routine.
FIG. 7 is a flowchart showing a dependency test processing routine in the embodiment.
FIG. 8 is an explanatory diagram illustrating an example of a dependency dictionary in the embodiment.
FIG. 9 is an explanatory diagram showing how dependency is established in a plurality of phrases.
FIG. 10 is an explanatory diagram showing priorities in dependency testing.
FIG. 11 is an explanatory diagram showing types when there are a plurality of dependencies in one input character string.
FIG. 12 is a flowchart showing a dependency test range determination process routine of the second embodiment.
FIG. 13 is an explanatory diagram showing an example sentence of a dependency test range determination process in the second embodiment.
FIG. 14 is an explanatory diagram illustrating an example of a dependency test order within a range in which dependency test is performed;
[Explanation of symbols]
21 ... CPU
22 ... ROM
23 ... RAM
24 ... Keyboard
25 ... Keyboard interface
26 ... CRT
27 ... CRTC
28 ... Printer
29 ... Printer interface
30: Hard disk controller (HDC)
31 ... Bus
32 ... Hard disk
40 ... Character input part
42. Conversion control unit
44 ... converted character string output section
50 ... Character string input part
52 ... Character storage
54 ... Independent word candidate creation section
56 ... Independent word analysis position management department
58 ... Autonomous dictionary
64 ... Attached word candidate creation part
66 ... Attached word analysis position management section
68 ... Attached word dictionary
70. Dependency learning department
70 ... Learning Department
72 ... Independent language learning department
74 ... Auxiliary Language Learning Department
76 ... Affix Learning Department
78 ... Character conversion learning unit
80: Word data creation unit
82 ... Connection verification department
84 ... Connection verification table
86 ... Cost calculator
90 ... Dependency candidate adjustment section
92 ... Passive analysis section
94 ... Participant tolerance analysis part
96 ... Dependency range management department
98 ... Dependency dictionary
99: Dependency transposition information adjustment unit
100: Word data storage unit
102 ... sentence segment writing section
104 ... Dependency transposition information adjustment unit
106 ... phrase data storage unit
108 ... Conversion character string output section

Claims (8)

A kana-kanji conversion device for inputting a kana character string and referring to a grammar dictionary to generate a kana-kanji mixed character string candidate,
A segmentation means for segmenting the input kana character string with reference to the grammar dictionary;
A dependency information dictionary storing dependency information between predetermined phrases;
Dependency verification means for verifying the presence of a dependency with other clauses in order from the one having the smallest distance between phrases, starting from the one phrase that has been divided, and referring to the dependency information dictionary;
Delimiter search means for searching for delimiters present in the input kana character string;
The clause that the delimiter is found, and a dependency test completion means for the end position of dependency test, further,
The dependency checking means starts from one of the divided clauses, and in the range to the end position of the dependency check,
Next candidate test means for sequentially determining the next candidate of the target phrase, with the test of the dependency test as a starting word for the phrase,
After the test by the next candidate test means is completed, the next candidate test means for continuing the test of the received word with the next candidate of the one phrase as the starting point as a related word;
Kanji conversion device provided with. A kana-kanji conversion device for inputting a kana character string and referring to a grammar dictionary to generate a kana-kanji mixed character string candidate,
A segmentation means for segmenting the input kana character string with reference to the grammar dictionary;
A range search means for searching a range delimited by a pair of codes indicating start and end in the input kana character string;
A dependency information dictionary storing dependency information between predetermined phrases;
Dependency verification means for verifying the presence of a dependency with other clauses in order from the one having the smallest distance between phrases, starting from the one phrase that has been divided, and referring to the dependency information dictionary;
A range searched by the range search means is excluded from a range of dependency verification by the dependency verification means, and a test range setting means for determining a dependency test range ; and
The dependency verification means includes the dependency verification range.
Next candidate test means for sequentially determining the next candidate of the target phrase, with the test of the dependency test as a starting word for the phrase,
After the test by the next candidate test means is completed, the next candidate test means for continuing the test of the received word with the next candidate of the one phrase as the starting point as a related word;
Kanji conversion device provided with. A kana-kanji conversion device according to claim 1 or 2,
  The segmentation means considers that the segment to be segmented is composed of independent words or independent words and ancillary words, and performs segmentation by a combination that minimizes the number of independent words constituting the input kana character string. Is a means to do
  Kana-kanji conversion device. A kana-kanji conversion device according to claim 1 or 2, wherein
The dependency verification means, when the segmented phrase is 3 or more, the one segment segmented as a starting point, the dependency information dictionary is referred to the presence of dependency with other phrases, of the three or more clauses, from near the end of the sentence side, a means you sequentially assayed temporary name kanji conversion device. A kana-kanji conversion device according to claim 4,
The dependency verification means is:
A first test means for first performing a test starting from a related word close to the end of the sentence;
When a dependency is found by the first verification means, pay attention to the phrase ahead of the phrase that is the starting point, and only the dependency that is established with the dependency found by the first verification means. A kana-kanji conversion device comprising: a second verification means that is taken out as a result of receiving. A kana-kanji conversion method for inputting a kana character string and referring to a grammar dictionary to generate a kana-kanji mixed character string candidate,
The input kana character string is segmented with reference to the grammar dictionary,
Starting from one of the above-mentioned clauses, refer to the dependency information dictionary storing the dependency information between the predetermined clauses, and determine the presence of dependency with other clauses with a small distance between clauses And test sequentially from
Search for a delimiter that exists in the input kana character string,
A clause that was found before Symbol delimiters, and end position of dependency test,
Furthermore, in the dependency test, in the range from one of the segmented clauses to the end position of the dependency test,
One phrase as the starting point is used as a dependency word, and the dependency test of the dependency test is sequentially performed on the next candidate of the target phrase.
A kana-kanji conversion method in which, after the test for the next candidate is completed, the test of the received word is continued using the next candidate of the one phrase as the starting point as a related word . A kana-kanji conversion method for inputting a kana character string and referring to a grammar dictionary to generate a kana-kanji mixed character string candidate,
In the input kana character string, a range delimited by a pair of codes indicating start and end is searched, and
The input kana character string is segmented with reference to the grammar dictionary,
With reference to the dependency information dictionary storing the dependency information of predetermined phrases while excluding the searched range from the dependency test range, starting from the one phrase that has been divided, and other The existence of the dependency with the clauses of the sentence shall be examined in order from the one with the smallest distance between the clauses ,
Furthermore, in the dependency test,
One phrase as the starting point is used as a dependency word, and the dependency test of the dependency test is sequentially performed on the next candidate of the target phrase.
A kana-kanji conversion method in which, after the test for the next candidate is completed, the test of the received word is continued using the next candidate of the one phrase as the starting point as a related word . A kana-kanji conversion method according to claim 6 or 7,
In the dependency test, when there are three or more divided clauses, the dependency information dictionary storing dependency information between predetermined clauses is used as a starting point. the presence of dependency with other clauses, of the at least three clauses, from the side close to the end of the sentence, you sequentially test
Kana-kanji conversion method.

Images