JPH0239366A - Kana/kanji conversion processing system - Google Patents

Abstract

PURPOSE:To improve KANA (Japanese syllabary)/KANJI (Chinese character) converting efficiency by selecting the KANJI candidates based on the connecting relation between the meaning element information on the paragraph undergone the KANJI conversion right before the due time point and the KANJI meaning element of the paragraph kept under the KANJI conversion. CONSTITUTION:In the case of input of a word 'YAKUIN SENKYO' (KANJI) consisting of two paragraphs, 'YAKUIN' (KANA) is inputted through an input device 11 and stored in an input buffer 12. Then the 'YAKUIN' (KANA) is collated with the header information contained in a dictionary 13 with this 'YAKUIN' used as an index key with push of a KANA/KANJI conversion instruction key. Thus the KANJI information 'YAKUIN' is obtained and at the same time the meaning element information 'SHITO' (KANJI) stored in the dictionary 13 is stored in a memory 16. Then a 2nd paragraph 'SENKYO' (KANA) is inputted and the KANA/KANJI conversion is instructed. Thus other 'SENKYO, SENKYO' (KANJI) are obtained out of the dictionary 13. Then 'YAKUIN' (KANJI) is compared with 'SHITO' (KANJI) and the first 'SENKYO' (KANJI) is defined as the first KANA/KANJI conversion candidate since the connection meaning information of the KANJI candidate shows 'SENKYO' equal to 'BASHO' and 'SENKYO' equal to 'SHITO' respectively.

Description

[Detailed Description of the Invention] [Field of Industrial Application] The present invention relates to a method for converting Japanese characters input in kana characters into sentences containing kanji and kana. This paper relates to a kana-kanji conversion processing method that efficiently converts kana-mixed sentences into kana-mixed sentences.

[Conventional technology]

FIG. 3 shows a schematic configuration of the conventional kana-kanji conversion method. Input device 11 is a device for inputting kana character strings.

The input buffer 12 is a device for temporarily storing input kana character strings, the dictionary 13 is a device for storing the correspondence between kana character strings and kanji character strings, and the output buffer 14 is a device for storing converted kanji character strings. A device to keep it.

The display device 15 is a device for displaying input kana character strings, converted kanji character strings, and the kana-kanji conversion process, and the control device 17 is a device for controlling the entire kana-kanji conversion process. A control device 17 that controls the whole stores the kana character string input from the input device 11 in an input buffer 12, and compares it with a dictionary 13 that stores the correspondence between kana character strings and kanji character strings, and selects the most probable character string. Kanji information was output to the output buffer 14, and the contents of the output buffer 14 were displayed on the display device 15.

[Problem that the invention seeks to solve]

In the above conventional technology, when performing kana-kanji conversion of an input kana character string, the semantic information of the character string that has already been converted to kanji is not used, so a large number of kanji candidates are output, and the kana-kanji conversion takes time. There was a drawback. The semantic information referred to here can be roughly divided into 0, ``person,'' ``
``thing'', ``thing (meaning a service, etc.)'', ``place'',
This is information representing "time" etc. The semantic information can be further classified and used as necessary. For example, the above-mentioned "object" can be subdivided and used at the level of "tools", "clothes", etc. Even in the past,
A compound clause-bracket conversion method is known in which a kana character string consisting of a plurality of clauses is input and then converted into a kanji character string in brackets. In the case of the compound phrase-bracket conversion method, a method is adopted in which a kana character string consisting of the multiple phrases input once is divided into phrase units, and the relationship of semantic elements between each phrase is converted into an optimal kanji string. has been done. However, a method that uses the semantic elements of a kanji string that has already been converted to convert the next input kana character string has not been adopted, which has the disadvantage that a large number of kanji candidates are output and it takes time to convert kana to kanji. there were.

The purpose of the present invention is to reduce the number of kanji candidates,
The object of the present invention is to provide a kana-kanji conversion processing method that greatly improves the efficiency of kana-kanji conversion.

[Means and effects for solving the problem] The kana-kanji conversion processing method of the present invention includes an input device for inputting a kana character string;
An input buffer that temporarily stores input character strings, a dictionary that stores the correspondence between kana character strings and kanji character strings together with the meaning information of the kanji character strings, and an output cover that stores converted kanji character strings. A display device that displays the input kana character string and the converted kanji character string and the kanji conversion process;
It consists of a storage device that stores the Kanji character string for one segment most recently converted into Kana-Kanji, that is, the semantic information of the clause immediately before the segment to be converted into Kana-Kanji, and a control device that controls the entire Kana-Kanji conversion process. be done. Then, the control device stores the kana character string input from the input device in the input cover sofa, matches the input kana character string with the kana character string in the dictionary, and when the two kana character strings match.

The candidate kanji character string paired with the matched kana character string in the dictionary is read out along with its semantic information. Next, this read semantic information is compared with the semantic information stored in the storage device, that is, the meaning information of the previous clause that has already been converted into kanji information, and the most closely related meaning is compared. Candidate kanji character strings with manufacturing information are used as solutions for kana conversion.

〔Example〕

An embodiment of the present invention will be described in detail below with reference to the drawings.

FIG. 1 is a block diagram of an embodiment of the present invention, showing an input device 11 for inputting a kana character string. An input cover sofa 12 that temporarily stores input kana character strings; a dictionary 13 that stores the correspondence between kana character strings and kanji character strings together with semantic information of the kanji character strings; An output buffer 14 for storing converted kanji character strings, a display device 15 for displaying input kana character strings, converted kanji character strings, and the kanji conversion process; Storage device 16 for storing semantic information of kanji information. and a control device 17 that controls the entire kana-kanji conversion process.

An example in which the present invention is applied to general kana-kanji conversion will be described below. Now, let's consider inputting the word "officer election," which consists of two clauses. FIG. 2 is a diagram showing the contents of the dictionary 13. When the first phrase "Yakuin" is input from the input device 11, the control device 17 stores the information of "Yakuin" in the input cover sofa 12 and transfers the input information to the output cover sofa 14.
and displays "Yakuin" on the second display device 15.

Next, the operator presses the "kana-kanji conversion instruction key" from the input device 11 in order to convert "yakuin" into kanji information. When the control device 17 receives the cough "Kana-Kanji conversion instruction key", it compares it with the kana heading information in the dictionary 13 using "Yakuin" in the input cover sofa 12 as an index key, and matches the input information "Yakuin". Kanji information with headings “
Officer”. Next, the control device 17
The part corresponding to "Yakuin" in the middle is the corresponding kanji information "
Convert to "Officer". Further, the control device 17 outputs the same information as the output buffer 14, that is, "Officer", as the display value W15, instead of the information "Yakuin" that has been displayed on the display device 15 up to now.

At the same time, the control device 17 stores, in the storage device 16, the meaning information ``person'' stored in the dictionary 13. Controller 17 then clears input buffer 12.

When the operator inputs the second phrase "senkyo" from the input device 11, the control device 17 stores "senkyo" in the input buffer 12 as before.

Next, the control device 17 stores "Senkyo" after "Officer" stored in the exit sofa 14.

That is, the control device 17 stores the information in the output sofa 14 in the form of "executive information". Next, the control device 17 outputs to the display device 15 the same information as the information in the output buffer 14, ie, “Jakugu Senkyo” instead of the information “Officer” displayed on the first display device 15. Next, when the operator presses the "kana-kanji conversion instruction key", the control device 17 checks the kana heading information of the dictionary 13 based on the information "senkyo" stored in the cover sofa 12, and Kanji information with a heading that matches the input information “Senkyo” is “
Obtain 2 of ``Occupy 1 Election''. Next, the control device 17
It is obtained that the conjunctive semantic information of the kanji candidates obtained by referring to 3 is occupied - place 1 election - person, respectively.
The connected semantic information of both candidate kanji and the phrase that has just completed kana-kanji conversion stored in the storage device 16, that is, ``
A comparison is made between the meaning information for ``officer'' and the meaning information for ``person,'' and the kanji ``election,'' which has the lowest meaning information, is selected as the first kana-kanji conversion candidate. Next, the control device 17 changes the content of the output buffer 14 from ""Officer Election"" to "Officer Election" and outputs the converted content of the output buffer 14, that is, "Officer Election" to the display device 15. do. At the same time, the control device 17 clears the semantic information "person" of "officer" in the storage device 16 and replaces it with the semantic information "thing" of "election".
A kana-kanji conversion processing method characterized by storing the information of "senkyo" in the input cover sofa 12 and clearing the information of "senkyo".

In this example, we explained the input of a word consisting of two clauses, but it is easy to apply the single-kana-kanji conversion processing method recursively, and it can also be applied to words composed of many clauses and general sentences. is self-evident.

〔Effect of the invention〕

As explained above, according to the present invention, unnecessary kanji can be removed by looking at the connection relationship between the semantic information of the clause that was just converted to kanji and the semantic element of the kanji of the clause that is currently undergoing kana-kanji conversion. No candidates are selected, and the efficiency of kana-kanji conversion can be greatly improved.

[Brief explanation of the drawing]

Fig. 1 is a block diagram of an embodiment of the kana-kanji conversion processing method of the present invention, Fig. 2 is a diagram showing the contents of a dictionary in the present invention, and Fig. 3 is a block diagram of an embodiment of the conventional kana-kanji conversion method. It shows. In the figure, 11 is an input device, 12 is an input cover, 13 is a dictionary, 14 is an output cover, and 15 is a display device. 16 represents a storage device, and 17 represents a control device. Patent applicant Nippon Telegraph and Telephone Corporation

Claims (1)

[Claims] A means for inputting Japanese kana character strings, a means for storing inputted kana character strings, a dictionary for associating kana character strings with kanji character strings, and a means for storing converted kanji character strings; In the kana-kanji conversion processing method, which has means for displaying the conversion process from a converted kanji character string or an input kana character string to a kanji character string, and a control means for controlling the entire kana-kanji conversion process, the kana character string and A means for storing the meanings of a kanji string in association with a kanji character string and the meaning of the previous clause to which the kanji string is connected, and a method for storing the meaning of a kanji string that has already been converted into kanji. When converting an input kana character string to a kanji character string, it is equipped with a means for storing semantics for one sentence segment, and when converting an input kana character string to a kanji character string, it is equipped with a means for storing the meanings of the kanji candidates of the input kana character string and the meanings that have already been converted to a kanji character string. A kana-kanji conversion processing method that compares semantic element information of a character string for one sentence segment immediately before a sentence, and displays and outputs a kanji candidate having the most closely related semantic element.