JPH0831098B2 - Document creation device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION (A) Field of Industrial Application The present invention relates to a document creation apparatus for creating a sentence containing kanji and kana from a reading input as input data.

(B) Conventional technology Readings such as kana and romaji are generally input from a keyboard as input data for converting a sentence containing kanji and kana. When the input pronunciation is instructed to be converted to a kanji / kana mixed sentence, the conversion part retrieves the kanji corresponding to the reading from the kanji dictionary memory to obtain the kanji and create the desired kanji / kana mixed sentence. To be done. For example, considering "school teacher" as a mixed kanji and kana sentence to be obtained after conversion, if "gakkou no sensei" is input as input data and then kanji conversion is instructed, the desired mixed kanji and kana sentence Is created (Japanese word processor SWP-340 manufactured by Sanyo Electric Co., Ltd., etc.).

However, in the above input data, after inputting "gakko" in the above input data, and giving a kanji conversion instruction, the kanji character string "school"
Even if you input “No Sensei” and instruct Kanji conversion, you cannot retrieve the appropriate Kanji character string or Kanji Kana character string starting with “No” from the dictionary, and the desired “No teacher”
Can't get

Therefore, in the above-mentioned prior art, a kanji kana mixed sentence containing independent words such as nouns, adjectives and verbs such as "school" and "teacher" and auxiliary words such as particles and auxiliary verbs such as "no" attached to these independent words. Regarding the conversion of, when starting the Kanji conversion, it is required that the first reading always starts with an independent word.
That is, the operator is aware of independent words and adjuncts, and in the case of "independent words + adjuncts ...", the Kanji conversion instruction key, and in the case of "adjuncts + independent words + adjuncts ..." It is necessary to perform multiple operations such as outputting the part as a kana character string, then inputting a kana character string starting from an independent word and converting into kanji.

On the other hand, if all adjuncts that are always expected when converting Kanji are checked, there is no need to input them from independent words.

However, this method not only takes time for checking, but also has a risk of incorrect conversion.
For example, if you read "Noyama" to get the main body, type "Noyama", ignore the relationship with the previous word, and check the possible auxiliary words, you will see It becomes an erroneous conversion.

(C) Problem to be Solved by the Invention The present invention lacks operability when inputting to always start from an independent word while being aware of attached words, as described above, and regarding all the attached words to the reading of the head part If the check is performed, not only will the conversion time be long, but there is also the possibility of erroneous conversion, and even if an adjunct that cannot be resolved by both of these comes to the beginning, there is a high probability that the desired kanji-kana mixed sentence will be generated. It is intended to provide the obtained character creation device.

(D) Means for Solving the Problem The present invention provides a storage device (57) for storing grammatical information of a phrase converted immediately before in a document creation device for inputting a reading and converting it into a kana-kanji mixed sentence. After this conversion, it is determined based on the grammatical information of the storage means (57) whether or not at least the beginning part of the newly input reading can be connected as an adjunct word to the phrase converted immediately before. And a judging means (58).

(E) Action In the present invention, since the storage means (57) stores the grammatical information of the phrase just converted, the judging means (58) uses this grammatical information to read at least the newly input reading. It is possible to determine whether or not to connect the phrase whose head portion is converted immediately before as an adjunct word.

(F) Embodiment FIG. 1 is a block diagram of an embodiment of the document creating apparatus of the present invention. (1) is connected by a signal bus (2) described later,
This is a main control unit that performs overall control by transmitting and receiving signals, and is configured by a microcomputer, and controls various controls according to a control program stored in advance in a built-in program memory. Reference numeral (3) is a keyboard as an input for inputting reading data, etc., a character key (31) for inputting the reading data, a conversion instruction key (32) for instructing a Kanji conversion process for reading, and a diagram are shown. Not various function selection keys,
Includes cursor movement keys, etc. (4) is an input control unit including a circuit for controlling the keyboard, and the keyboard (3) is a signal bus (2) via the input control unit (4).
Connected to.

Reference numeral (5) is a Kanji conversion unit for converting the reading input from the keyboard (3) into a Kanji or Kanji / Kana mixed sentence under the control of the main control unit (1), and a conversion control unit (51) for controlling the conversion process. , A dictionary part (52) in which kanji data, kana data, alphanumeric data, etc. are stored based on JIS code, a dictionary search part (53) for searching the dictionary part (52) based on the reading instructed to convert, As a result of the search, a candidate table (54) for storing information such as frequency, character string length, and part-of-speech, etc. of candidates of the corresponding kanji / kana mixed sentence, the above candidate table (54)
A candidate determination unit (55) that outputs the candidate with the highest priority based on the information of (1), and a previous phrase determination processing unit that determines the converted output as the previous phrase when the operator selects and confirms the candidate currently output ( 56), a pre-clause buffer (57) for storing grammatical information such as the determined part-of-speech and presence or absence of annex
And the grammatical information stored in the preceding clause buffer (57), and an adjunct word determining unit (judges whether or not at least the leading portion of the next input reading data is connectable to the preceding clause as an adjunct word ( 58). (6)
Is an output unit such as a CRT display, an LCD display, and a plasma display that sequentially displays conversion candidates and confirmed Kanji-mixed sentences during input reading and Kanji conversion processing.

Next, a concrete example will be described with reference to the transition of the display screen of FIG. 2 based on the input of "gakko / no-sensei /" (/ is a kanji conversion instruction) and the kanji conversion instruction. In FIG. 2, the underlined (un) part means that the kanji conversion process is to be performed, and (ca) indicates a cursor.

Use the character keys (31) on the keyboard (3) to first
As shown in FIG. 5A, the reading "gakko" is input, and the kanji conversion instruction key (31) is operated for the reading to instruct the kanji conversion. Based on such an instruction, the main control (1) activates the kanji conversion unit (5), the dictionary search unit (53) executes the search of the dictionary unit (52), and the word group corresponding to the reading, for example, "image , I, school, individual, high, school ”are extracted, and information such as frequency and character string length part of speech is stored in the candidate table (54) using these word groups as candidates. Since the preceding clause buffer (57) does not store anything in the preceding clause buffer (57) by default, the adjunct word determination clause (58) is not determined here and the candidate determination section (55) is determined.
On the basis of the information stored in the candidate table (54), the longest method and part-of-speech check are performed, and the candidate "school" having the highest priority is output to the output unit (6) as shown in FIG. 2 (b). .

Here, the operator inputs the next reading "NO SENSEI" from the character keys (31) in order to confirm the "school" currently output and displayed (Fig. 2 (c)). As a result, the preceding phrase "school" is confirmed, the preceding phrase determination processing unit (56) is activated, and its contents "school" and "part of speech: noun, no auxiliary word"
Is stored in the previous clause buffer (57).

The Kanji conversion instruction key (32) is operated for the above-mentioned "No Sensei" to activate the Kanji conversion unit (5). The dictionary search unit (53) searches the dictionary unit (52) and stores a candidate word group “field, put, put on, teacher, preemptive, ...” As the search result in the candidate table (54) together with grammatical information. . Next, the adjunct word determination unit (58) is activated to take out the grammar information of the preceding phrase buffer (57) in which the phrase converted immediately before, that is, "school, part of speech: noun, no adjunct word" is stored. The adjunct word determination unit (58) that has taken out "noun, no adjunct word" judges whether at least the beginning part of the reading "no-sensei" currently input based on the grammatical information is attached as an adjunct word. The determination unit (58) determines that the leading reading "no" is an adjunct word as a particle, and stores "no (no independent word part, adjunct word) in the candidate table (54).

Finally, the candidate determination part (55) is activated and the two phrases of “no (no independent word, adjunct word) teacher (part of speech: noun, no adjunct word)” are decided for the reading “nosensei”. , Outputs the sentence "Kono-sensei" mixed with Kanji and Kana in the output section. That is, the display screen becomes “school teacher” as shown in FIG.

When the "no teacher" is confirmed, the previous clause buffer (5
7) has been cleared and the preceding clause "Teacher, part of speech: noun,
"No attached word" is stored and prepared for the next Kanji conversion process.

As described above, in the present embodiment, the input means for inputting the reading as the input data to be converted into the kanji / kana mixed sentence, the conversion instructing means for instructing the conversion of the above reading into the kanji / kana mixed sentence, and the conversion instructing means. And a conversion means for converting to a kanji / kana mixed sentence based on the instruction of
Storage means for storing the grammatical information of the phrase that has been converted immediately before and has already been determined, and whether or not at least the leading part of the input reading has been connected as an adjunct word to the phrase that has been converted immediately before. If the result of the determination by the determination means is that connection to the immediately preceding converted phrase is possible, a kana character string is output to the connectable reading part. ing.

(G) Effect of the Invention In the present invention, since the storage means (57) stores the grammatical information of the phrase converted immediately before, the judging means (58) at least reads the reading newly input by this grammatical information. Since it is possible to determine whether or not the beginning part can be connected as an adjunct to the phrase converted immediately before, it is possible to check the adjunct in consideration of the relationship with the preceding word.

[Brief description of drawings]

FIG. 1 is a block diagram of a document creating apparatus according to the present invention, and FIG.
Drawing (a) -Drawing 2 (d) are mimetic diagrams showing a change of a display screen. (1) ... Main control unit, (3) keyboard, (5) ... Kanji conversion unit, (6) ... Output unit, (31) ... Character key, (32) ... Conversion instruction key, (57) ... Previous clause buffer , (58)… Adjunct word judgment section.

 ─────────────────────────────────────────────────── ─── Continued Front Page (72) Inventor Masahiro Ishiba 2-18 Keihan Hondori, Moriguchi City, Osaka Sanyo Electric Co., Ltd. (56) Reference JP-A-61-173377 (JP, A)

Claims (1)

[Claims] 1. A document creating device for inputting a phonetic reading and converting it into a kanji kana mixed sentence, a storage means (57) for storing the grammatical information of the phrase converted immediately before, and a new input after this conversion. Determining means (58) for determining, based on the grammatical information of the storage means (57), whether at least the beginning part of the reading is connectable as an adjunct word to the phrase converted immediately before. Characterized document creation device.