JPH07129571A - Document creating apparatus and document creating method - Google Patents

Abstract

PURPOSE:To convert a Japanese reading into an English index through the use of the dictionary of a small storage capacity by dividing the Japanese reading into units such as syllables to be Japanese reading parts and retrieving a parts dictionary registering Japanese reading parts and English index parts based on it. CONSTITUTION:A character string conversion part 3 the Japanese reading into the English index but when the English index corresponding to the Japanese reading is not registered in a general dictionary 4, the part 3 outputs the Japanese reading to a parts conversion part 5. The parts conversion part 5 divides the received Japanese reading such as the syllables, part elements by each character, etc., for example, retrieves the parts dictionary 6 in accordance with the character string of the divided Japanese reading and extracts the English index parts. Furthermore, the part 5 synthesizes the extracted English index parts and prepares plural synthetic indexes. The character string conversion part 3 retrieves an English index word registered in the general dictionary 4 in accordance with the plural synthetic index candidates and extracts the English index coincident with the synthetic index.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a document creating apparatus and a document creating method for converting input Japanese reading data into English headings.

[0002]

2. Description of the Related Art In a conventional text creating apparatus, for example, when converting Japanese reading into English heading, the following processing is performed. Using the Japanese reading data as input data, search the dictionary in which Japanese reading and corresponding English headings are registered (hereinafter referred to as general dictionary), and input characters and Japanese reading of the general dictionary are all characters. By inputting the corresponding English heading when the columns match, the input Japanese reading is converted into an English heading.

However, when an English heading is written in Japanese reading, the writing method in Japanese may differ depending on the individual. For example, reading "computer" as "konpyuta" or "konpyuta", but the same English heading, but different readings in Japanese. Therefore, when inputting Japanese readings, data corresponding to a plurality of readings is registered in the conventional general dictionary so that even if the readings are slightly different, they can be converted into the same English heading. In the above example, the data "computer" = "konpyuta" and "computer" = "konpyuta" are registered respectively so that the input Japanese reading can be converted to either of the two. ing. in this way,
For reading Japanese, I used a dictionary that registered multiple English headings.

[0004]

When an English heading is represented by Japanese reading in this way, many Japanese readings for the same English heading occur due to individual differences. For this reason, the conventional document creation apparatus requires a dictionary that covers a large number of different Japanese readings for the same English heading. However, if English headings are registered in the dictionary so that there is a one-to-one correspondence with a large number of different Japanese readings, in addition to the large storage capacity required to store the dictionary, the input Japanese readings If is not registered in the dictionary, there was a problem that had to be solved, such as the Japanese reading could not be converted into an English heading.

SUMMARY OF THE INVENTION It is an object of the present invention to solve the above problems and to provide a document creating apparatus and a document creating method capable of converting a Japanese reading into an English heading by using a dictionary having a small recording capacity.

[0006]

In order to achieve the above object, in the first aspect of the present invention, at least one headword for the reading is set based on the input reading input by the input means for inputting the reading. The stored word is retrieved, and the entry word for the input reading input from the input means is extracted by the extraction means.

As a result of the search processing by the means for extracting the entry word, if the entry word cannot be extracted,
The input reading is decomposed into parts elements, and the reading parts are obtained by the reading decomposition means for obtaining the reading parts.

Next, the means for storing the entry word parts for the reading parts obtained by decomposing the reading into parts elements are searched,
The entry word component is extracted by the means for extracting the entry word component for the reading component obtained from the reading decomposition unit.

The headword parts extracted by the means for extracting the headword parts are combined, and a combined headword is obtained by the reading and combining means for obtaining a combined headword. At least 1 for reading based on this synthetic headword
The means for storing one entry word is searched, and the entry word that matches the synthetic entry word is extracted from the means for storing the entry word.

In the second aspect of the invention, the input reading is input by the input means for inputting the reading, the input reading input from the input means is decomposed into the component elements, and the reading decomposition means for obtaining the read component is used. Get the parts.

Next, the means for storing the entry word parts for the reading parts obtained by decomposing the reading into parts elements are searched,
The entry word component is extracted by the means for extracting the entry word component for the input reading component decomposed by the reading decomposition unit.

A synthesis headword is obtained by a synthesis means for synthesizing the entry word components extracted by the means for extracting the entry word component to obtain a synthesis headword. Search for a means that stores the entry word based on this composite entry word,
The headword is extracted by the means for extracting the headword that matches the synthetic headword.

The third aspect of the present invention is whether or not all the headword parts constituting the composite headword, which are synthesized by the headword synthesizing means, are represented by an alphabetic character string of one character. To judge. When the determination means determines that the combined entry word is an alphabetic combined entry word, the combined entry word is set as one of the entry word candidates for the input reading.

[0014]

According to the structure of the present invention, when the Japanese reading is not registered in the general dictionary, the Japanese reading is divided into units such as syllables to make the Japanese reading parts, and the disassembled Japanese reading parts. Since the Japanese reading can be converted to the English heading by searching the parts dictionary that stores the Japanese reading parts and the English heading parts based on can do.

[0015]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT An embodiment of the present invention will be described with reference to FIG. In the figure, a keyboard 1 is for inputting a character string such as Japanese reading. The keyboard 1 has a key for inputting characters such as kana, numbers and symbols, a key for instructing conversion of a character string, a key for selecting and determining the converted character string, etc. Is provided.

The input control processing unit 2 identifies what key is input from the keyboard 1. Character string converter 3
Converts the Japanese reading, which is the character code string obtained by the input control processing unit 2, into an English heading. The character string conversion unit 3 first searches a general dictionary 4 described later to determine whether or not Japanese reading is registered in the general dictionary 4. If it is determined that the corresponding English headline is registered, it is taken out from the general dictionary 4. If the English heading for Japanese reading is not registered in the general dictionary 4, the Japanese reading is output to the component conversion unit 5.

The general dictionary 4 is a dictionary in which a character code string of Japanese reading and a character code string of an English heading corresponding to the Japanese reading are registered, and Japanese reading is registered in kana or the like. However, the English headings in the general dictionary 4 are registered corresponding to the most common Japanese readings.

The component conversion unit 5 decomposes the Japanese reading received from the character string conversion unit 3 into component elements such as syllables and characters. According to the decomposed character string of Japanese reading (hereinafter referred to as Japanese reading component), the component dictionary 6 described later is searched, and the English heading component corresponding to the Japanese reading component is extracted. Further, the extracted English headline parts are combined to create a composite headline.

In the parts dictionary 6, Japanese reading parts and corresponding English heading parts are registered. Also, with respect to one Japanese reading part, a plurality of candidates are registered as English heading parts in consideration of spellings that are not pronounced in Japanese.

The display 7 is, for example, a 20-line liquid crystal display. Next, the operation of the embodiment will be described with reference to the flowchart shown in FIG. In the present embodiment, an example will be described in which the Japanese reading consisting of a kana character code string "sentebu" is converted into an English heading "sensitive".

From the keyboard 1, the Japanese reading "Senshibu" is keyed in (step S1). The input control processing unit 2 discriminates the data input by the key and inputs the character code string data such as Japanese reading, or the conversion key etc. for converting Japanese reading into English heading. To determine if. When the input data is Japanese reading, the Japanese reading is output to the character string conversion unit 3. When the reading headline conversion key is input, the process proceeds to the character string conversion unit 3. The character string conversion unit 3
The general dictionary 4 is searched with the Japanese reading of "SENSEBU" entered by key input (step S2). The general dictionary 4 is shown in FIG.
As shown in (a), Japanese reading and corresponding English headings are stored. The character string conversion unit 3 determines whether or not the Japanese reading input in the general dictionary 4 is stored (step S3). In this example, since the Japanese reading "senseitebu" is not registered in the general dictionary 4, it is determined that there is no English heading candidate. Then, the processing shifts to the component conversion unit 5. The component conversion unit 5 decomposes the input Japanese reading into a plurality of Japanese reading components (step S4). On the other hand, the parts dictionary 6 is searched using the Japanese reading parts, and if the corresponding Japanese reading parts are registered in the parts dictionary 6, the corresponding English heading parts are acquired (step S
5). As shown in FIG. 4B, the parts dictionary 6 stores Japanese reading parts and corresponding English heading parts. When disassembling the Japanese reading parts input here into the Japanese reading parts, the same reading as the Japanese reading parts registered in the parts dictionary 6 is performed.

In this example, the input Japanese reading "sentebu" is converted into "se""n""shi""te""bu" or "sen""shi""te""bu" etc. Disassemble, and from the parts dictionary 6, “se” = “se”, “n” = “n” “m”,
"Shi" = "shi""si","te" = "ti",
"Bu" = "bu""ve","sen" = "sen""t"
Search for candidates for English headline parts such as "hen".

Subsequently, the parts converting section 5 combines the candidates for the English headline parts to create a combined headline (step S).
6). At this time, for example, "x" and "x" may not be adjacent to each other, and a combination rule of parts may be applied to prevent improper combination. In this example, (1) 's
enshitibu ”, (2)“ sensitiv ”
e ”, ... (n)“ Sensitive ”, and the like can be obtained as candidates for a plurality of combined headings. In this way, the component conversion unit 5 moves the processing to the character string conversion unit 3 after synthesizing the English headline components and obtaining a plurality of synthesized headings. The character string conversion unit 3 searches for an English heading registered in the general dictionary 4 according to a plurality of composite heading candidates, and extracts an English heading that matches the composite heading (step S).
7). Then, the extracted English headline is displayed on the display 7 (step S8). If there are a plurality of matching entry word candidates, for example, the entry word candidates obtained by searching the general dictionary 4 are stored in the entry candidate buffer (not shown), and the entry word is displayed on the display 7. Display one candidate. Further, when displaying the candidate of the entry word, if there is a candidate of the entry word other than the displayed entry word, for example, by pressing the “next candidate” key (not shown) for displaying the candidate of another entry word. It will be shown that by pressing or the like, even undisplayed headword candidates can be displayed and selected. Alternatively, as another method, a list of candidates may be displayed on the display 7 as homonyms and synonyms so that a synthetic heading can be selected from the list. In this example, the nth “se
Since "native" is registered in the general dictionary 4, "sensitiv" is used as an English headline from the general dictionary 4.
“E” is extracted, and the English headline “sensitive” obtained as a result of this conversion is displayed on the display 7.

As a second embodiment, in the above embodiment, the Japanese is read by the component conversion unit 5 from the beginning without searching the general dictionary 4 from the input Japanese reading (steps S2 and S3 in the above embodiment). It is also possible to start the process of decomposing the reading.

Further, in this case, only the English headings for conversion are registered in the general dictionary 4, and the parts conversion unit 5 combines the candidates of the English heading parts to create a composite English heading. The English headline that matches the headline is extracted from the general dictionary 4.

As a third embodiment, as in the above two embodiments, the parts conversion unit 5 uses the parts dictionary 6 to convert the Japanese reading parts into English heading parts and synthesizes these English heading parts. Combine to a composite headline. Then, after that, whether or not the English headline parts that are the component elements that form the composite headline obtained by combining are composite headlines (hereinafter referred to as alphabetic composite headlines) all represented by a character string of one letter of the alphabet. Make a decision. If it is determined that all English headline parts are alphabetic composite headlines represented by a character string of one alphabet,
The general dictionary 4 is not searched for the alphabet composite headline, and the alphabet composite headline is displayed on the display 7 as a candidate for the English headline for the input Japanese reading. When the alphabet composite headline obtained in this way is displayed on the display 7, the characters may be displayed in English capital letters.

The flowchart of FIG. 3 shows this operation. In the same flowchart, the same parts as those in FIG. In FIG. 3, the component conversion unit 5 searches the component dictionary 6 for a Japanese reading component to obtain an English heading component (step S5). Next, the English headline parts are combined to create a combined headline (step S
6) It is determined whether or not the obtained combined heading is an alphabetic combined heading (step S9). At this time,
The number of English headline parts in this composite headline is compared with the number of character strings in the composite headline, and if they match, it is determined to be an alphabetic composite headline. When there is a candidate corresponding to the alphabet composite headline in this way, the general dictionary 4 is not searched based on the alphabet composite headline, and the alphabet composite headline is displayed on the display 7 (step S8). On the other hand, if there is no corresponding candidate, the general dictionary 4 is searched based on the synthetic heading, and the English heading matching the synthetic heading is extracted (step S).
7). Then, the extracted English headline is displayed on the display 7. Further, when the general dictionary 4 is searched based on the synthetic heading and the matching English heading is found, the synthetic heading may be displayed on the display 7 without extracting the English heading from the general dictionary 4. Further, as described in the second embodiment, the input Japanese reading can be directly divided to obtain the English heading without performing the processes of steps S2 and S3 in FIG.

Here, the case where the Japanese reading "Je-Efuke-" is converted into the English heading "jfk" will be described as an example. General dictionary 4 "Je-Efuke-" =
If "jfk" is not registered, the Japanese reading is decomposed and the parts dictionary 6 is searched. In this example, the input Japanese reading "Je-Efuke-" is decomposed by the component conversion unit 5 into "Je-", "Efu", "Ke-" and the like. From the Japanese reading parts obtained by disassembling with these parts conversion unit 5, the parts dictionary 5
Search for "Je-" = "j", "jey", "Efu" =
English candidate parts such as "f", "ef", "ke-" = "k", "key" are obtained. Next, the candidates for the English headline parts are combined (1) “jeyfkey”, (2) “j
eyfk ”, (3)“ jfkey ”, (4)“ jfk ”
A synthetic heading candidate such as ... Is obtained. In this example, since the English headline parts are "Jee-", "Efu", and "Ke-", the composite headline is composed of three English headline parts. From this, it is determined that the number of character strings is three, and the third candidate is represented by the alphabetic character string is the fourth candidate “j”.
Since it is "fk", it is judged that this combined heading corresponds to the alphabetic combined heading and is displayed on the display 7. On the other hand, for the synthetic headings other than the fourth candidate “jfk”, the English headings registered in the general dictionary 4 are searched, and the matching English heading is selected, and the selected one is displayed as the homonym of the alphabetic synthetic heading “jfk”. Display on 7.

In addition to the method described above, the following method may be used to determine whether the composite headline is an alphabetic composite headline. The input Japanese reading "Je-Efuke-" is disassembled by the parts conversion unit 5, and "Je-", "Efu",
Set it to “ke-” etc. Next, the parts dictionary 5 is searched from these Japanese reading parts, and "je-" = "j""jey","ef" = "f""ef","ke-" = "k""key". , Etc.) to obtain candidates for an English heading part. At this time, "Je-" =
If there is a candidate represented by one letter in all the English heading parts such as “j”, “Efu” = “f”, “Ke-” = “k”, it is represented by this one letter. The English headline parts are combined to create an alphabetic composite headline and displayed on the display 7.

As in the above embodiment, a parts dictionary is searched from the Japanese reading parts obtained by decomposing the input Japanese readings to create a plurality of composite headline candidates, and then the general dictionary is searched. By doing so, even if the input Japanese reading is not registered in the general dictionary, it can be converted into an English heading. Therefore, according to the present invention, it is not necessary to register all the English headings corresponding to Japanese reading as in the conventional case, only by adding the parts dictionary in which the Japanese reading parts and the corresponding English heading parts are registered. It is possible to reduce the total storage capacity required for the dictionary, including the dictionary and the parts dictionary.

[0031]

As described in detail above, according to the present invention, even a Japanese reading not registered in the general dictionary can be converted into an English heading by using the parts dictionary. In other words, it is not necessary to cover all the Japanese readings in the general dictionary, so that the storage capacity of the general dictionary can be reduced. Therefore, even for dictionaries including the general dictionary and the parts dictionary, the parts dictionary can be commonly referred to in the disassembled reading, so that all the necessary storage capacity of the entire dictionary can be reduced.

[Brief description of drawings]

FIG. 1 is a block diagram showing the configuration of a document creation device according to a first embodiment.

FIG. 2 is a flowchart showing the operation of the embodiment shown in FIG.

FIG. 3 is a flowchart showing the operation of the third embodiment.

FIG. 4 is a general dictionary 4 and a parts dictionary 6 according to the first embodiment.
FIG.

[Description of Codes] 1 ... Keyboard, 2 ... Input control processing unit, 3 ... Character string conversion unit 4 ... General dictionary, 5 ... Parts conversion unit, 6 ... Parts dictionary,
7 ... Display.

Claims (6)

[Claims] 1. An input means for inputting a reading, a means for storing at least one headword for the reading, a means for storing at least one headword for the reading, and input from the input means. When the headword cannot be extracted as a result of the search processing performed by the means for extracting the entry word for the input reading and the means for extracting the entry word, the input reading is decomposed into component elements, and the reading is performed. The reading disassembling means for obtaining the parts, the means for storing the entry word parts, and the means for storing the entry word parts for the reading parts obtained by disassembling this reading into the component elements are obtained from the reading disassembling means. , A means for extracting the entry word component for the reading component, and a synthesis for combining the entry word components extracted by the means for extracting the entry word component, The apparatus further comprises headword synthesizing means for obtaining a headword, and means for extracting the headword that matches the synthesized headword synthesized by the reading synthesizing means from the means that stores at least one headword for the reading. A document creation device characterized by the above. 2. An input means for inputting a reading, a reading decomposing means for decomposing the input reading input from the input means into component elements to obtain a reading component, and a reading component decomposing the reading into component elements. Means for storing the entry word component for, and a means for searching the means for storing the entry word component to extract the entry word component for the reading component decomposed by the reading decomposition means; By combining the headword parts extracted by the extracting means, a headline combining means for obtaining a combined headword, a means for storing the headword, and a means for storing the headword, the headword combining is performed. And a means for extracting the entry word that is obtained by combining in the means and matches the combined entry word. 3. Whether or not all of the headword parts constituting the composite headword that are composed by the headword composing means are alphabetic composite headwords represented by an alphabetic character string of one character. And a means for setting the composite headword as one of the headword candidates for the input reading when the composite headword is judged to be an alphabetic composite headword by the judgment means. The document creating apparatus according to claim 1, which is provided. 4. An entry word for the input reading information is extracted from dictionary information storing at least one entry word for reading based on the input reading information, and the entry word is extracted as a result of the process of extracting the entry word. When it cannot be extracted, the input reading information is decomposed into parts elements to be a reading part, and based on the reading parts that decompose the input reading information into parts elements, a heading for the reading parts that decomposes the reading into parts elements. An entry word component for the reading component is extracted from the dictionary information that stores the word component, then the extracted entry word components are combined to create a combined entry word, and the combined entry word obtained is combined. Based on the dictionary information storing at least one headword for the reading, the headword matching the synthetic headword is extracted. How to create a document. 5. The reading component is obtained from dictionary information in which input reading information is decomposed into component elements to be a reading component, and based on the decomposed reading component, an entry word component for the reading component obtained by dividing the reading into component elements is stored. After extracting the entry word parts for the above, extracting the entry word parts, synthesizing the entry word parts to create a combined entry word, and storing the entry words based on the combined entry words obtained by combining A method of creating a document, characterized in that the entry word that matches the synthetic entry word is extracted from dictionary information. 6. It is determined whether or not the headword parts forming the composite headword obtained by combining are alphabetic composite headwords all represented by an alphabetic character string of one character, and the combining is performed. The document according to any one of claims 4 and 5, wherein when the entry word is determined to be an alphabet composite entry word, the composite entry word is set as one of the entry word candidates for the input reading information. How to make.