JP2575947B2 - Phrase extraction device - Google Patents

Description

The present invention relates to a phrase extracting apparatus, and more particularly, to a phrase extracting apparatus suitable for a language processing device such as a word processor, a translation device, a proofreading device, and a device using a database. The present invention relates to a phrase extracting device.

(B) Conventional technology Conventionally, in order to extract a phrase for translation or proofreading from a converted Japanese document into which the information of the phrase delimiter is not inserted, an independent word dictionary, an auxiliary word dictionary, an affix dictionary, etc. Dictionaries and a table showing the connection relationship of those elements,
The segmentation was performed while referring to tables such as a grammar table.

(C) Problems to be Solved by the Invention In such a conventional phrase extraction apparatus using dictionaries and tables, (1) a large amount of storage devices are required to store the dictionaries and tables. Become.
(2) The time required for the phrase extraction processing for performing a dictionary search or a table search becomes longer. (3) There is a problem that a control program for segment extraction becomes complicated. Also,
In phrase extraction processing, a device that extracts a phrase by determining the character type of each character in a character string has also been considered. However, in such a device, (4) the phrase is cut out mechanically according to the character type, so that a phrase with a lot of kana characters may be cut incorrectly. In other words, Japanese orthography has not been established and spread, so even the same words are often written in kanji or hiragana, and the current tendency is that hiragana writing is increasing, I often cut phrases.

The present invention has been made in view of the above circumstances, and provides a phrase extracting apparatus that can solve the above problem.

(D) Means for Solving the Problems FIG. 1 is a block diagram showing the basic configuration of the present invention. Referring to FIG. 1, the present invention provides a sentence storing means 1 for storing a sentence, an instructing means 2 for giving an instruction for reading out a desired sentence from the sentence storing means 1, and a reading for reading out a designated sentence from the sentence storing means 1. Means 3, character type discriminating means 4 for discriminating a character type such as kanji, hiragana, katakana, etc., one character at a time from the read sentence and discriminating a punctuation mark; Means 5 and a phrase extracting means 6 for inserting a break in a read sentence when a predetermined character type transition point and a punctuation mark are detected for the character type stored in the discrimination result storage means 5.
A hiragana cut-out dictionary means 7 for storing characters or character strings that should be accompanied by a break in terms of continuous hiragana characters, and a counting means 8 for counting the number of continuous hiragana characters as a result of character type determination. When the count value of the counting means 8 satisfies a predetermined condition, it is compared with the characters or character strings stored in the hiragana cut-out dictionary means 7 and the discrimination result is stored for each of the verified characters or character strings. Hiragana extraction means 9 for extracting a phrase from the Hiragana character string stored in the means, storage means 10 for storing a sentence output from the Hiragana extraction means 9, and an output for visually outputting the sentence stored in the storage means 10 This is a phrase segmentation device including the means 11.

The phrase extraction device according to the present invention can be applied to a word processor, a translation device, a proofreading device, a device using a database, and the like. Further, a phrase extraction process is also required in a device that outputs a sentence by voice. Can be applied.

(E) Function According to the present invention, when a unit of the cutout process, for example, one minute is cut out from the text stored in the text storage device 1, the character type of each character constituting the sentence is determined by the character type determining means 4. And the determination result is sequentially stored in the determination result storage means 5. Next, the phrase extracting means 6 inserts a break in the character string at the change of, for example, a kana to kanji character, and temporarily stores the result. The number of the hiragana characters to be counted is counted, and if a predetermined condition is satisfied, the character is compared with the character or character string for the hiragana extraction stored in the hiragana extraction dictionary means 7. The phrase extraction position is corrected so that a break in the phrase is made at a position after the character or the character string, and the obtained result is output to the output means 11 so as to be visually output.

(F) Embodiment The present invention will be described in detail below based on an embodiment shown in the drawings.
The present invention is not limited by this.

FIG. 2 is a block diagram showing an embodiment in which the present invention is applied to a Japanese word processor. In the figure, reference numeral 20 denotes a word processor. Reference numeral 21 denotes a sentence / text storage device, such as a proppy disk or a hard disk as an external storage device, or a RAM or the like as an internal storage device.
Alternatively, it can be constituted by, for example, a database or the like as another storage device, and stores a Japanese document composed of kana-kanji mixed sentences.

Reference numeral 22 denotes a keyboard as an instruction means for editing text,
It is provided with character input keys for performing sentence proofreading, various kinds of instruction keys, and the like, and inputs a sentence and an instruction for reading out a desired sentence from the sentence storage device 21. twenty three
Is a reading device that cooperates with the CPU 24, and reads a desired sentence specified by the keyboard 22 from the text storage device 21. twenty five
Is a character type discrimination device that cooperates with the CPU 24, and is a sentence storage device 2
For the sentence read from 1, character types such as kanji, hiragana, katakana, etc. are determined one by one, and punctuation marks are determined. Reference numeral 26 denotes a result storage device serving as a determination result storage means and a storage means, which is constituted by a RAM and has a character type determination device 25.
Are sequentially stored, and the sentence cut by the later-described phrase extraction device is stored.
The phrase extraction device 27 cooperates with the CPU 24 to generate the result storage device 26.
Based on the inflection points and punctuation marks of the character types stored in the text, a break is inserted in the read sentence. Reference numeral 28 denotes a Hiragana cut-out dictionary that stores characters or character strings that should be accompanied by a break in the phrase for consecutive Hiragana characters, and is composed of a ROM. 29 is a hiragana cutout processing device as a counting means and a hiragana cutout means, cooperates with the CPU 24, counts the number of consecutive hiragana of the character type determination result,
When the count value satisfies a predetermined condition, it is compared with the characters or character strings stored in the Hiragana extraction dictionary 28, and for each of the verified characters or character strings, the Hiragana stored in the result storage device 26 is stored. Extracts a phrase from a character string. Reference numeral 30 denotes a display device as an output means connected to the CPU 24 via the output control unit 31. The display device 30 is composed of a dot matrix type display device such as a CRT or an LCD, and displays a sentence with a break between phrases.

In such a configuration, the phrase segmentation process according to the embodiment will be described with reference to the flowchart shown in FIG. 14 using the example sentence "Materials to be submitted to the JPO."
FIG. 3 shows a state where, for example, one sentence is cut out from the sentence storage device 21 in which a sentence file and other databases are stored in accordance with the processing unit, and the sentence is stored in the result storage device 26.

As described above, when a desired sentence is read from the sentence storage device 21 according to the processing unit, for example, one sentence, one paragraph, one chapter, etc. (step 60), the character type of the sentence is determined. The code is determined (step 61). For more information,
Since each character of the read sentence is assigned a unique character code such as JIS code, the character type is compared with each condition of the code determination table shown in FIG. Determine. In the conditions, cc is the character whose character type is to be determined, a1 and b1 represent the beginning and end of the Kanji code, a2 and b2 represent the beginning and end of the Hiragana code, and a3 and b3 represent the beginning and end of the Katakana code. A4 represents a period.

FIG. 5 shows the result of determining the character type of the example sentence. here,
“Kan”, “hi”, “ka”, and “phrase” are symbols indicating a kanji code, a hiragana code, a katakana code, and a period code, respectively. This determination result is stored in the result storage device.
It is stored in 26 (step 62).

Next, the contents of the stored character type codes are determined by referring to the determination points in the extraction determination table shown in FIG. 6 (step 63). This table shows conditions for extracting a phrase in accordance with the arrangement of character type codes and operations to be performed when the conditions are satisfied. That is, (1) a keyword break (segment break) is made at a transition point from Hiragana to Kanji. (2) Make a keyword break at the transition point from Hiragana to Katakana.
(3) Make a keyword break after the period.

Next, a cutout symbol, for example, "/" is inserted at the end of the phrase according to the processing of the cutout determination table, and the result is stored in the result device 26 (step 64). FIG. 7 shows the result obtained by the above-described phrase extraction processing. If the word “made” is written as “made”, the character type discrimination result shown in FIG.
Although the phrase can be cut correctly using the cut-out determination table shown in the figure, in the example sentence that is currently being processed, "made" is written in hiragana, Have failed. Today, when the ratio of kanji in text is decreasing, there are many sentences in Hiragana notation like this example sentence.

Next, the processing of the hiragana cutout processing device 29 will be described. FIG. 8 shows a buffer for counting the continuation of the hiragana character type shown in FIG. 5 (hereinafter abbreviated as HL). For explanation, the character type determination results of the input example sentences shown in FIG. 5 are shown side by side. . FIG. 9 shows a rule for dividing a hiragana character string into phrases according to the value of HL. FIG. 10 is a diagram for explaining the conditions shown in FIG. Here, S is the start position of the kana character string, and E is the end position. Therefore, E
−S = HL. M represents a marker for cutting a phrase.
The marker may have a plurality of characters. In this case, the position of the last character is M. EM = MAL and the marker M
Indicates the value of the character string after kana.

FIG. 11 and FIG. 12 show the contents of the Hiragana extraction dictionary 28 to be collated at the time of Hiragana phrase extraction, and are a set of Kana character string extraction characters or character strings. FIG. 11 is a set of one to four characters to be collated and having a break at the end of the character. In FIG. 12, the number of characters to be collated is two, and a break is inserted between the characters (between "and" and "i"). That is, in a character string such as "", "", "", and the like, a segment break is inserted between "" and "". Note that the kana cutout characters and character strings shown in FIGS. 11 and 12 are classified in the order of ordinary character codes.

From the character type discrimination result in FIG. 5 or in parallel with obtaining the character type discrimination result, the number of continuous characters in a continuous kana character string is counted (step 65), and as shown in FIG. Store in the Hiragana continuous counting buffer. The number of consecutive characters in Hiragana for "I made it." Therefore, by referring to the kana extraction condition shown in FIG. 9 (step 66), it is found that the condition HL ≧ 4 is satisfied. Then, a character or character string for extracting a hiragana character string shown in FIG. 11 is referred to (step 67). In the case of example sentences, the first character "" can be matched. (Step 68).

In addition, it can be seen that MAL ≧ 3 from the position of “」 ”, and thus it can be seen that all the conditions shown in FIG. 9 are satisfied. Thus, the condition shown in FIG. 9 is satisfied,
In addition, since the character matches the kana cutout character in FIG. 11, a hiragana cutting operation is executed according to the rule in FIG. 9 (step 69). As a result, the phrase position can be corrected by separating the hiragana character string "was created." As "wa / was created." (Step 7)
0). FIG. 13 shows the segmentation result obtained by the above processing.

Next, if the end condition is no, that is, if there is a next sentence to be segmented, the next sentence is stored in the text storage device 21.
Is performed, and control is performed so that the storage positions do not overlap (step 71).

If yes in step 71, that is, if there is no longer any object to be cut out, the necessary information is stored in the result storage device.
The data is stored in 26, and the termination is processed (step 72).

(G) Effects of the Invention According to the present invention, (1) when extracting a phrase, a dictionary is not used, so that the configuration of the apparatus can be simplified.
Thus, the present invention can be applied not only to word processors and office computers, but also to other small devices, specifically, electronic notebooks and calculators with program functions. (2) The phrase extraction processing and the keyword search processing can be performed at high speed. (3) A control program for segment extraction can be simplified.
(4) When extracting a phrase, there is a drawback that the position of the phrase cannot be known because Japanese is written in solid, so it takes a lot of processing and time to determine where to start and end the phrase. I have. According to the present invention, the language processing can be performed after the phrase position is determined, so that the processing time can be significantly reduced.
(5) It is possible to improve the segmentation accuracy of a sentence with a large number of hiragana notations, which has conventionally caused many segmentation mistakes.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a basic configuration of the present invention, FIG. 2 is a block diagram showing a configuration of a word processor according to an embodiment of the present invention, and FIG. 3 shows an example of sentences stored in a text storage device. FIG. 4 is an explanatory diagram showing the contents of a character type determination table, FIG. 5 is an explanatory diagram showing a character type determination result,
FIG. 6 is an explanatory diagram showing the contents of the cutout determination table, FIG.
FIG. 8 is an explanatory diagram showing a cutout result, FIG. 8 is an explanatory diagram showing the contents of a hiragana continuous counting buffer, FIGS. 9 and 10 are explanatory diagrams showing a hiragana character string cutting rule, FIG. 11 and FIG. FIG. 13 is an explanatory diagram showing a hiragana cutout character, FIG. 13 is an explanatory diagram showing a modified phrase cutout result according to the embodiment, FIG.
FIG. 14 is a flowchart for explaining the operation of the embodiment. 1 ... text storage means, 2 ... instruction means, 3 ... readout means, 4 ... character type determination means, 5 ... determination result storage means, 6
...... phrase extraction means, 7 ... hiragana extraction dictionary means,
8 ... counting means, 9 ... hiragana extraction means, 10 ... storage means, 11 ... output means.

Claims (1)

(57) [Claims] 1. A sentence storing means for storing a sentence, an instructing means for giving an instruction for reading a desired sentence from the sentence storing means, a reading means for reading a designated sentence from the sentence storing means, and a read sentence Character type discriminating means for discriminating the character type such as kanji, hiragana, katakana, etc., one by one and discriminating punctuation points; discrimination result accumulating means for sequentially accumulating the discrimination results by the character type discriminating means; When a predetermined character type transition point and a punctuation mark are detected for a character type, a phrase cutout means for inserting a break in the read sentence, and for a continuous hiragana character, a character or a character string that should have a break in the phrase. Hiragana cut-out dictionary means stored, counting means for counting the consecutive number of hiragana of character type discrimination result, and count value by counting means When a predetermined condition is satisfied, the character string or character string stored in the Hiragana cut-out dictionary means is collated, and for each character or character string collated, the Hiragana character string stored in the discrimination result storage means. A phrase extracting apparatus comprising: hiragana extracting means for extracting a sentence from the phrase; storage means for storing a sentence output from the hiragana extracting means; and output means for visually outputting the sentence stored in the storing means.