JPH03282874A - Document preparation device - Google Patents

Abstract

PURPOSE:To always obtain a correct conversion result by deriving automatically exact part-of-speech information corresponding to registered words and phrases in connection with an input of the registered words and phrases and a reading character- string being registration objects, and registering it together with the registered words and phrases and the reading character-string. CONSTITUTION:When an input of registered words and phrases and a reading character- string is finished, a CPU 1 executes an analytic processing of a part of speech corresponding to the registered words and phrases, based on a part-of-speech analytic program stored in a memory 2. That is, the CPU 1 decides whether the registered words and phrases are the predicates or the substantives, based on part-of-speech analytic data stored in a table 2c of the memory 2 from the registered words and phrases and the reading character-string. Subsequently, when part-of-speech information corresponding to the registered words and phrases is determined, the CPU 1 executes a processing for registering the inputted registered words and phrases and the reading character-string and the determined partof-speech information in a new word dictionary memory 3, based on a new word registering program stored in the memory 2. In such a way, at the time of KANJI (Chinese character) conversion processing, the registered words and phrases of a correct part of speech are outputted, and a correct conversion result is always obtained.

Description

DETAILED DESCRIPTION OF THE INVENTION [Object of the Invention (Field of Industrial Application) The present invention relates to a document creation device having a new word registration function.

(Prior Art) Conventionally, document creation devices have a new word registration method as a function to ensure that necessary words are obtained as a result of the kanji conversion when input kana characters are converted into kana-kanji characters to create a document. This new word registration method is a method in which a specified registered word (new word) is registered in the new word registration dictionary along with its corresponding reading character string, and when the reading character string is input after registration, the corresponding registered word is output as a kanji conversion result. It is.

By the way, the new word registration method has the function of inputting part-of-speech information corresponding to a registered word and outputting the registered word as a kanji conversion result based on the corresponding part-of-speech information. There is. With this method, even if the input character strings have the same pronunciation, for example, for words used as verbs, registered words registered as nouns can be prevented from being output as conversion results. be able to.

(Problems to be Solved by the Invention) As a new word registration method, there is a method of inputting part-of-speech information corresponding to a registered word or phrase. However, when inputting part-of-speech information, there is a problem in that incorrect part-of-speech information is input manually, and a registered phrase is not output as a proper conversion result.

SUMMARY OF THE INVENTION An object of the present invention is to provide a document creation device that can register accurate part-of-speech information corresponding to registered words and realize a reliable new registration function in a new word registration system.

[Configuration of the Invention (Means and Effects for Solving the Problems) The present invention provides a document creation device having a new word registration function, which uses registered words and phrases input based on part-of-speech analysis data prepared in advance and their corresponding words. A device comprising a part-of-speech analysis means for analyzing the part of speech of a registered word from a pronunciation character string, and a registration means for registering the registered word, the pronunciation character string, and the part-of-speech information obtained by the part-of-speech analysis means in a new word registration dictionary. It is.

With this configuration, it is possible to obtain accurate part-of-speech information corresponding to the registered word as the registered word and its corresponding reading character string are input, and to realize a new word registration function using this part-of-speech information. Become.

(Example) The present invention will be described in detail below with reference to the drawings.

FIG. 1 is a block diagram showing the main parts of a document creation device according to the same embodiment. As shown in FIG. 1, this apparatus is provided with a microprocessor (CPU) 1 which is a control circuit that performs document creation processing and controls the entire apparatus. C
The PUI executes each program stored in the memory 2 to perform new word registration processing. The memory 2 has a storage area 2a that stores a new word registration control program, a part of speech analysis program, and a new word registration program.

It has 2b and 2d. Furthermore, the memory 2 has a table 2c that stores data for part-of-speech analysis.

Furthermore, this device includes a keyboard 4 for inputting various input data necessary for document creation processing, a display section 5 for displaying kana-kanji conversion results, created documents, etc., and registered words, reading character strings, parts of speech, etc. A new word dictionary memory 3 is provided for registering information.

Next, the operation of this embodiment will be explained.

First, when a necessary kana character string is input from the keyboard 4, the CPU performs a kana-kanji conversion process to convert the input kana character string into a kanji character string. The conversion result is displayed on the screen of the display unit 5. Here, when the next candidate key on the keyboard 4 is operated, a plurality of kanji conversion results corresponding to the same human-powered kana character string are output and displayed on the screen of the display unit 5.

When the selection/execution key 42 of the keyboard 4 is operated when the necessary kanji conversion result is displayed, the kanji conversion result is confirmed.

Here, when the new word registration key 41 of the keyboard 4 is operated in order to register the confirmed kanji conversion result as a new word (registered word/phrase), the CPU A registration control operation is executed, and the display unit 5 and keyboard 4 are set to the input mode for registered words (step Sl in FIG. 2). As a result, a message instructing the input of a registered word is displayed on the screen of the display unit 5, and when a registered word is specified by operating the cursor keys on the keyboard 4, the specified registered word (i.e., a new word to be registered) is displayed. A certain kanji conversion result) will be input. Next, the CPUI sets the display unit 5 and the keyboard 4 to an input mode for reading character strings corresponding to the registered words (step S2 in FIG. 2). As a result, a message instructing the manual effort of the pronunciation string is displayed on the screen of the display unit 5, and when the pronunciation string is input from the keyboard 4, the designated pronunciation string is input.

When you finish inputting the registered words and reading strings, the CPU
executes the part-of-speech analysis process corresponding to the registered words based on the part-of-speech analysis program stored in the memory 2 (step S3). That is, the CPU stores the registered words and reading character strings in the table 2c of the memory 2. Based on the data for part-of-speech analysis, the registered words and phrases are divided into words (verbs, adjectives,
Determine whether it is an adjective (verb) or a nominal (noun (9 pronouns)). Based on this judgment result, in the case of predicates,
The CPUI executes a part-of-speech analysis process for the predicate based on part-of-speech analysis data as shown in FIG. 3, for example (steps S4 and S5). On the other hand, if the judgment result is nominal, CP
U1 executes the part-of-speech analysis process for nominal speech based on the part-of-speech analysis data as shown in FIG. 4, for example (step S4
．． S8).

CPUI uses part-of-speech information obtained through part-of-speech analysis processing as 1.
It is determined whether or not there are different types, and if there are multiple types, a process for making a final decision is executed (step 86.S9). Specifically, multiple types of part-of-speech information are displayed on the screen of the display unit 5,
Instructs to select and specify by operating the cursor keys on the keyboard 4.

When the part-of-speech information corresponding to the registered word is determined in this way, the CPUI uses the input registered word, reading character string, and determined part-of-speech information based on the new word registration program stored in the memory 2. A process of registering the word in the new word dictionary memory 3 is performed (step 87).

After such new word registration is completed, when a kana character string including the reading character string corresponding to the registered word is input and the kanji conversion process is executed, an appropriate conversion result will be obtained based on the registered part of speech information. The character string corresponding to the reading character string will be converted into a registered word/phrase. In other words, accurate part-of-speech information corresponding to registered words is registered through part-of-speech analysis processing, so registered words with the correct part of speech are output during kanji conversion processing, making it possible to always obtain appropriate conversion results. becomes.

[Effects of the Invention] As detailed above, according to the present invention, in the new word registration method, accurate part-of-speech information corresponding to the registered word is automatically generated as the registered word to be registered and the reading character string are input. can be registered along with the registered word/phrase and reading character string. Therefore, part-of-speech information corresponding to registered words can always be accurately registered. This results in
When performing kanji conversion processing, it is possible to always obtain appropriate conversion results by using registered words and phrases that correspond to kana human character strings.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing the main parts of a document creation device according to an embodiment of the present invention, FIG. 2 is a flowchart for explaining the operation of the embodiment, and FIGS. 3 and 4 are respectively the same embodiment. FIG. 2 is a conceptual diagram for explaining part-of-speech analysis data related to . 1...CPU, 2...Memory, 3...New word dictionary memory, 4...Keyboard, 5...Display section.

Claims (1)

[Scope of Claims] A document creation device having a new word registration function, comprising an input means for respectively inputting a registered word to be registered as a new word and a pronunciation string corresponding to the registered word, and a part-of-speech analysis device prepared in advance. a part of speech analysis means for analyzing the part of speech of the registered word from the registered word and phrase inputted by the input means and the reading character string based on data; 1. A document creation device comprising: a registration means for registering part-of-speech information in a dictionary for new word registration.