JPH1031672A - Person name conversion device and method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To properly perform the person name conversion processing without constructing any table that contains a large number of KANJI (Chinese character) candidates by collating a synthesized person name with an exception dictionary and outputting a person name which is obtained by deleting the KANJI other than those used for the person names. SOLUTION: An analysis means 21 supplies the punctuation information used for dividing the KANA (Japanese syllabary) character string received from an input means 1 to a KANJI retrieval means 22 as a line of partial character strings. The means 22 sets the position information contained in the divided character strings as a retrieval flag. If a relevant single KANJI is included in a single KANJI dictionary storage part 25, the single KANJI where the retrieval flag is coincident with a stored position flag is stored as a candidate. Then a synthesis means 23 synthesizes the retrieved single KANJI with correspondence secured to the position flag. A collation means 24 retrieves a name inappropriate to reading among the synthesized candidates via an exception dictionary storage part 26. Then the means 24 collates the retrieved name with a synthesized name and excludes the coincident synthesized character string candidates. These candidates are displayed and selected by a user.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a Japanese word processor and a personal name conversion apparatus and method which function in a software program similar thereto, and in particular, automatically display kanji candidates when a personal name is read. .

[0002]

2. Description of the Related Art Conventionally, when a Japanese word processor or a similar software program functioning on a personal computer is used to convert input readings into kanji, kanji candidates for readings are sequentially or so-called by a so-called dictionary function. A list is displayed, and the user selects an appropriate kanji candidate from these candidates. In such a dictionary function, a kanji candidate corresponding to reading is selected with reference to a kanji candidate table stored in a memory.

The kanji candidate table generally stores proper nouns such as place names and personal names, in addition to general terms described in an ordinary dictionary. Also, when a long sentence is input, a corresponding kanji candidate can be displayed for each phrase by cooperating with a phrase determination function for dividing the long sentence into phrases.

[0004]

However, proper nouns appearing in sentences, especially personal names, have various readings and various candidates more than general nouns. It was necessary to build a Kanji candidate table.

[0005] For example, even with the same reading name "Kazuya", there are various kanji candidates such as "Kazuya", "Kazuya", "Kazuya", and "Kazuya". Is an inappropriate candidate. Therefore, in order to display all of these as kanji candidates, the dictionary must be registered in advance in the kanji candidate table, and the storage capacity of the memory for storing the kanji candidate table increases. I was

The present invention has been made in view of the above-described problems in the conventional personal name conversion processing, and has as its object the purpose of constructing this kind of kanji candidate table without constructing a huge kanji candidate table for personal names. It is an object of the present invention to provide a person name conversion device and method capable of appropriately performing a person name conversion process.

[0007]

The invention according to claim 1 of the present application is an analysis means for dividing a kana character string of a person's name input from an input means and analyzing the reading with the divided character string. A single kanji dictionary storing single kanji corresponding to the reading of the divided character string analyzed by the analysis means, an exception dictionary storing kanji other than the personal name notation, and a single kanji stored in the single kanji dictionary Search means for searching for a single kanji character corresponding to the divided character string from above, synthesizing means for synthesizing a personal name from the single kanji searched by the search means and the arrangement of the divided character strings, Matching means for matching a personal name with the exception dictionary and deleting kanji other than the personal name notation, and output means for outputting a personal name obtained by deleting the kanji other than the personal name notation. Character conversion equipment characterized by Located in.

According to the first aspect of the present invention, the reading is analyzed by the divided character string obtained by dividing the kana character string of the personal name input from the input means, and the divided character analyzed by the analyzing means is analyzed. A single kanji corresponding to the divided character string is searched from a single kanji stored in a single kanji dictionary that stores a single kanji corresponding to the reading of the column, and a personal name is obtained from the searched single kanji and the arrangement of the divided character strings. Are combined, and the combined personal name is compared with an exception dictionary that stores kanji other than personal name notation, and kanji other than personal name notation is deleted. As a result, the finally obtained personal name is output.

Further, the invention according to claim 2 of the present application sets a headword whose heading is a reading corresponding to the divided character string and a position in the divided character string where the headword is located. The single kanji dictionary configured to store the position information obtained and the single kanji set by the position information, and the single kanji corresponding to the divided character string is searched by the position information. Item 1 is a personal name conversion device according to item 1.

[0010] According to the second aspect of the present invention, in the configuration of the first aspect, the single kanji corresponding to the divided character string is a headword whose heading is a reading corresponding to the divided character string. Is searched based on the position information that sets which position of the divided character string is placed.

According to the invention described in claim 3 of the present application, the kana character string of the personal name inputted from the input means is divided, and the reading is analyzed by the analyzing means with the divided character string. The search unit searches the single kanji dictionary for a single kanji corresponding to the reading of the divided character string, and synthesizes a personal name from the single kanji searched by the search unit and the arrangement of the divided character strings. Compare the combined personal name with the exception dictionary and delete the kanji other than the personal name notation by the matching means,
A personal name conversion method characterized in that the collating means outputs a personal name obtained by deleting a kanji character other than the personal name notation by an output means.

Further, the invention according to claim 4 of the present application searches for a reading corresponding to the analyzed divided character string from a headword, and uses the position information of the headword to read the reading corresponding to the divided character string. 4. The personal name conversion method according to claim 3, wherein a single kanji is searched.

[0013] According to the above-mentioned inventions,
Without building a huge kanji candidate table for personal names,
This kind of personal name conversion processing can be appropriately performed.

[0014]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Preferred embodiments of the present invention will be described below in detail with reference to the drawings.

FIG. 1 is a block diagram showing one embodiment of a personal name conversion apparatus according to the present invention. As shown in FIG. 1, this personal name conversion device mainly includes an input means 1 such as a keyboard, a data processing device 2 operated by program control, a storage device 3 for storing information, and an output means 4 such as a display device. It is composed of

The input means 1 is for inputting readings in kana, romaji or the like. The data processing device 2 includes an analysis unit 21 that divides the kana character string corresponding to the reading input from the input unit 1 into partial character strings for convenience, and a single kanji dictionary to be described later based on the kana character string divided by the analysis unit 21. Retrieval means 2 for retrieving the corresponding single kanji by referring to storage unit 25
2, a synthesizing unit 23 for synthesizing the retrieved single kanji corresponding to the reading, and a matching unit 24 for removing an inappropriate kanji string from the synthesized kanji string by referring to an exception dictionary storage unit 26 described later. Contains.

The storage device 3 stores single Kanji characters frequently used in personal names and also stores a position flag indicating position information in reading a personal name in which the single Kanji characters may be used. A single kanji dictionary storage unit 25 and an exception dictionary storage unit storing a combination of single kanji characters that are not considered to be used for a certain reading.

The output means 4 displays on the display the reading input input from the input means 1 and the kanji candidate group processed by the data processing device 2 and searched.

Next, the details of one embodiment of the single kanji dictionary storage unit 25 and the exception dictionary storage unit 26 will be described with reference to FIGS.

FIG. 2 is a table showing an embodiment of dictionary information stored in the single kanji dictionary storage unit 25. This table includes a headword 311 and a position flag 312 indicating position information in the kanji string. And the kanji candidate 313 are stored in advance.

Here, the headword 311 indicates a reading corresponding to each of the kanji candidates, and the position flag 31
Numeral 2 indicates, for example, whether the reading is likely to be located at the head, in the middle, or at the end of the personal name. For example, the reading "Kazu" is likely to be located at the beginning of the name, such as "Kazuya", "Kazuko", "Kazuo", and at the end of the name, such as "Shigekazu", "Kamozu"there's a possibility that. On the other hand, it is rarely in the middle of the name. The position flag indicates such conventionally determined position information as three-digit information.
For example, in the case of “Kazu”, there is a possibility that it may be located at the beginning and end of the name, so the position flag is “101”.

The kanji candidate 313 has a one-to-many relationship with the headword 311, and the reading of the same “kazu” is “one”.
Single kanji such as "wa" correspond to each.

On the other hand, as shown in FIG. 3, the exception dictionary storage unit 26 associates a headword 321 with a personal name 322 that is inappropriate as a kanji string corresponding to the headword in a one-to-one or one-to-many relationship. It is a table.

Next, the operation of the personal name conversion apparatus according to the present invention will be described with reference to FIGS.

First, the analyzing means 21 adds delimiter information for dividing the kana character string input from the input means 1, and the kana character string to which the delimiter information is added is converted into a partial character string as a kanji search means. 22.

For example, when a kana character string “Kazuya” is input, the reading analysis means converts the kana character string such as “Kazuya”, “Ka / zuya”, “Kazu / ya”, and “ka / zu / ya”. To divide. Note that "/" indicates delimiter information,
A kana character or a kana character string delimited by the delimiter information is a partial character string.

Each of the divided character strings obtained by the analyzing means is converted into a single kanji dictionary storage unit 25 by the kanji searching means 22.
, The corresponding single kanji candidate is searched. FIG. 4 is a flowchart showing the processing performed by the search means 22.

First, position information of the divided character string divided by the analyzing means 21 in the entire reading is set as a search flag (step 401). At this time, for example, when trying to search for a divided character string "Kazu", "Kazu"
Since it is located at the beginning of the kana character string “Kazu / ya”, “100” is set as a position flag for search. Then, the pronunciation "Kazu" is set as a key item at the time of retrieval (step 402), and a headword and a single kanji corresponding to this key item are retrieved (step 403). If the corresponding headword and single kanji are stored in single kanji dictionary storage unit 2
5 is determined (step 404).
If the corresponding single kanji exists, it is further determined whether the search flag matches the position flag stored in the single kanji dictionary storage unit 25 (step 405). If all of these determinations are satisfied, the retrieved single kanji is stored as one of the candidates (step 406).

The processing from step 403 to step 406 is repeated until all the partial characters and the single Kanji corresponding to the search flag are searched from the single Kanji dictionary storage unit 25. As a result, for example, for the partial character “Kazu”, “1” and “Japanese” are searched, and for the partial character “Ya”, “ya”, “ya”, “ya”, “arrow”, “eight”, etc. Single Kanji candidates are searched.

If a reading "Yaichi" is input and is divided into partial character strings, it becomes "Yaichi", "Ya / ichi", "Ya / chi", "Ya / i / chi". However, if an attempt is made to search for a single kanji character string corresponding to “yai / chi”, there is no single kanji candidate for “yai”, and thus an incorrect single kanji candidate is not searched. When searching for a single kanji character string corresponding to “ya / ichi”, the search flag is set to “100” because “ya” is located at the head, but a single kanji candidate having the same headword “ya” is searched for. Of these, since "ya" and "ya" do not match the search flag, these single kanji candidates are not searched.

When these processes are completed, the retrieved single kanji is sent to the next combining means 23.

The synthesizing means 23 synthesizes the searched single kanji in association with the position flag. For example, "Kazu / ya"
In the case of, from the combination of single Kanji candidates searched earlier, Combined by means of synthesis.

Next, in order to exclude an inappropriate name from the candidates combined by the combining means 23, the exception matching means 2
4 and the exception dictionary storage unit 26 are used. FIG. 5 is a flowchart showing the processing performed by the matching unit 24.
First, the matching unit 24 searches the exception dictionary storage unit 26 for an inappropriate name for the given reading (step 501), and compares the name combined by the combining unit 23 with the searched inappropriate name. (Step 502), and it is determined whether or not the name matches the inappropriate name (step 503). If they match, the combined character string candidate is excluded (step 504). In the case of this embodiment,
Since the kanji character string "wahachi" is inappropriate for the reading "kazuya" in the exception dictionary storage unit 26, "wahachi" is removed from the synthesized kanji character string. You.

As described above, the candidates obtained through the collating means 24 are a group of candidates for the appropriate personal name for the input reading, and these are displayed sequentially or collectively and wait for selection by the user.

[0035]

As is apparent from the above description, according to the present invention, the kana character string of the personal name input from the input means is divided, and the reading is analyzed by the divided character string.
A single kanji corresponding to the read of the analyzed divided character string is searched from the single kanji dictionary, a personal name is synthesized from the searched single kanji and the arrangement of the divided character strings, and the synthesized personal name is used as an exception dictionary. By matching, the kanji other than the personal name notation is deleted, and the personal name obtained by deleting the kanji other than the personal name notation is output.As a result, without building a huge kanji candidate table for the personal name, This kind of personal name conversion processing can be appropriately performed.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an embodiment of a personal name conversion apparatus according to the present invention.

FIG. 2 is a diagram for explaining the inside of a single kanji dictionary storage unit 25 in the personal name conversion device of FIG.

FIG. 3 is an exception dictionary storage unit 3 in the personal name conversion device of FIG.
It is a head for explaining the inside of 2.

4 is a kanji search means 22 in the personal name conversion device of FIG.
6 is a flowchart showing the processing performed by the first embodiment.

FIG. 5 is a flowchart showing a process performed by an exception collation processing unit 24 in the personal name conversion device of FIG. 1;

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Input means 2 Data processing device 3 Storage device 4 Output means 21 Analysis means 22 Search means 23 Synthesis means 24 Exception matching means 25 Single Kanji dictionary storage part 26 Exception dictionary storage part

Claims (4)

[Claims] An analysis means for dividing a kana character string of a person's name input from an input means and analyzing the reading with the divided character string, and corresponding to the reading of the divided character string analyzed by the analysis means A single kanji dictionary for storing single kanji characters, an exception dictionary for storing kanji characters other than the personal name notation, and a search unit for searching a single kanji character corresponding to the divided character string from the single kanji stored in the single kanji dictionary A synthesizing means for synthesizing a person's name from the single kanji searched by the searching means and the arrangement of the divided character strings, and comparing the personal name synthesized by the synthesizing means with the exception dictionary to obtain kanji other than the personal name notation. A personal name conversion device, comprising: matching means for deleting; and output means for outputting a personal name obtained by deleting a kanji character other than the personal name notation. 2. A headword whose heading is a reading corresponding to the divided character string, position information that sets where the headword is placed in the divided character string, and the position information. 2. The single kanji dictionary configured to store the single kanji and the single kanji corresponding to the divided character string is searched by the position information.
The person name conversion device according to 1. 3. A kana character string of a person's name input from an input means, and the analyzing means analyzes the reading with the divided character string, and a single kanji corresponding to the reading of the analyzed divided character string. From the single kanji dictionary to the search means, and the synthesizing means synthesizes a personal name from the single kanji searched by the search means and the arrangement of the divided character strings, and compares the synthesized personal name with the exception dictionary to A personal name conversion method, wherein a kanji other than a personal name notation is deleted by a collating means, and the collating means outputs a personal name obtained by deleting the kanji other than the personal name notation, to an output means. 4. The method according to claim 1, wherein a reading corresponding to the analyzed divided character string is searched from a headword, and the single kanji corresponding to the divided character string is searched based on position information of the headword. 3. The personal name conversion method according to 3.