JPH01219947A - Automatic word learning device - Google Patents

Abstract

PURPOSE:To improve the conversion efficiency at the time of converting KANJI (Chinese character) by learning automatically a word being insufficient in a dictionary and registering it in the dictionary. CONSTITUTION:Reading which is to be formed to a KANJI (Chinese character) and KANA (Japanese syllabary) mixed sentence is inputted from a reading input part 11, and sent out to a KANA to KANJI converting part 12 and a reading input storage part 15. The KANA to KANJI converting part 12 executes a KANA to KANJI conversion by using a dictionary 13 and stores it in a notation storage part 14. When a word corresponding to reading does not exist in the dictionary 13, the KANA to KANJI converting part 12 retrieves a one- character KANJI by its sound or Japanese reading of a Chinese character and sends out a retrieved one-character KANJI train to a word learning part 16. The word learning part 16 generates one table in which reading and the notation are allowed to correspond to one-to-one from reading of the reading input storage part 15, and a notation code train and a reading train of the one-character KANJI train which is sent out of the KANA to KANJI converting part 12. An automatic registering part 17 registers one table which has been generated by the word learning part 16, in the dictionary 13.

Description

DETAILED DESCRIPTION OF THE INVENTION Field of the Invention The present invention relates to an automatic word learning device used in Japanese word processors, personal computer beaters, terminal devices having Japanese input functions, and the like.

BACKGROUND OF THE INVENTION FIG. 2 shows the configuration of a portion of a conventional Japanese word processor for inputting kana and kanji.

In FIG. 2, reference numeral 21 denotes a reading input section, which inputs a reading character string to be converted into a kana-kanji conversion section 22. 23 is a dictionary prepared in advance for performing kana-kanji conversion. 24 is a notation storage unit that stores the kana-kanji conversion results. 25 is a user dictionary that can be stored independently by the user.

Next, the operation of the above conventional example will be explained. A reading (hiragana character string) to be converted into a sentence containing kana-kanji is inputted from the reading input section 21, and this reading is converted into kana-kanji in the kana-kanji conversion section 22.

As is known, Kana-Kanji conversion follows Japanese grammar, performs morphological analysis and syntactic analysis, performs a dictionary search using 23 dictionaries, converts into a kanji-kana mixed sentence, and the conversion result,
That is, the notation representing Kanji, Hiragana, Katakana, etc. in JIS code is stored in the notation storage unit 24. In this way, by converting from reading to kanji-kana mixed sentences,
Can input Japanese. If the words provided in the dictionary 23 are insufficient, a user dictionary 25 is provided in which the user inputs and memorizes the Yomiji notation, and a dictionary search using both the dictionary 23 and the user dictionary 25 allows conversion. This can make up for the lack of word count.

In this way, even with the above-mentioned conventional kana-kanji conversion device, by registering the missing words in the user dictionary, it is possible to make up for the shortage in the number of words.

Problems to be Solved by the Invention However, with the user dictionary in the conventional kana-kanji conversion device, the user extracts only words that do not exist in the dictionary, inputs the pronunciation and the corresponding notation, and registers them as words. There was a problem that when inputting concatenated hiragana character strings for kana-kanji conversion, words that did not exist in the dictionary could not be registered in the user dictionary. The present invention solves these conventional problems, and provides an excellent dictionary automatic learning device that can automatically register a word that does not exist in the dictionary in the dictionary when the word is read manually. The purpose is to

Means for Solving the Problems In order to achieve the above object, the present invention provides a reading input means, a kana-kanji conversion means, a dictionary, a notation storage means for storing the JIS code of the conversion result, a reading input storage means, and a kanji conversion means. A means for associating words with pronunciations, which are a set of It is designed to have a learning function so that it can be registered.

Effects The present invention has the following effects due to the above configuration. That is, the pronunciation for making a kanji-kana mixed sentence, that is, a hiragana character string, is input, and these readings are analyzed and converted into a kanji-kana mixed sentence by dictionary search.

If a word does not exist in the dictionary, it automatically learns the word that does not exist in the dictionary by converting it into a single kanji that corresponds to the pronunciation and registering a set of these - character kanji as a word in the dictionary. You will be able to fill in the missing words.

Embodiment FIG. 1 shows the configuration of an embodiment of the present invention. In FIG. 1, reference numeral 11 denotes a reading input section, which inputs the input reading 5 to the kana-kanji conversion section 12.

13 is a dictionary. The notation of the kanji-kana mixed sentence obtained by the kana-kanji conversion is stored in the 14-notation storage unit. 1
5 is a reading input storage unit that stores input readings independently of the kana-kanji conversion unit, and 16 is a learning unit that compiles sets of readings and kanji and makes correspondences between readings and words, and automatically registers the learned results. The unit 17 registers the information in the dictionary 13.

Next, the operation of the above embodiment will be explained. In the above embodiment, the reading input section 11 should be a mixed sentence of kanji and kana.
The reading input unit 11 sends the reading to the kana-kanji conversion unit 12 and the reading input storage unit 15. The kana-kanji conversion unit 12 performs morphological and syntactic analysis of the reading according to Japanese grammar, and performs a dictionary search using the dictionary 13. These are similar to known kana-kanji conversions. Here, if the word corresponding to the pronunciation does not exist in the dictionary 13, the kana-kanji conversion unit 12 searches for the one-character kanji by its sound or kun. The kana-kanji conversion unit 12 sends the searched one-character kanji string to the word learning unit 16. The word learning unit 16 generates a set of readings and notations in one-to-one correspondence from the readings in the reading input storage unit 15 and the notation code (JIS code) string of the one-character kanji string sent from the kana-kanji conversion unit 12 and the reading string. Create a table. The automatic registration section 17 registers a set of tables created by the word learning section 16 in the dictionary 13.

In this way, according to the above embodiment, when a pronunciation corresponding to a word that does not exist in the dictionary 13 is input, the kana-kanji converter 12 converts it into a one-character kanji string. By combining these kanji strings into words, combining the pronunciation and notation, and registering them in the dictionary 13, when a pronunciation of a word that does not exist in the dictionary 13 appears while the user is performing kana-kanji conversion, It has the advantage of being able to automatically learn and register words.

An example of operation is shown in FIG. 1 is a case where the target word exists in the dictionary. Cases 2 and 3 are cases in which there are no words for the pronunciations βdendo' and 'zofuku', respectively. 4 is an automatically learned word.

By automatically learning the word 4, even if the reading of 2°3 is input next, the same conversion as 1 can be performed.

Effects of the Invention As is clear from the above embodiment, the present invention automatically learns words that are missing in the dictionary and registers them in the dictionary, allowing the user to write sentences containing kanji and kana through kana-kanji conversion. This has the advantage that the device automatically learns the words used by the user. By performing kana-kanji conversion using the words used by the user, the conversion rate can be improved.

[Brief explanation of the drawing]

Figure 1 shows word automation 11 in one embodiment of the present invention.
・Reading input section, 12...Kana-kanji conversion section, 13...
Dictionary, 14... notation storage section, 15... reading input storage section, 16... word learning section, 17... automatic registration section. Name of agent: Patent attorney Toshio Nakao (1st person)
Figure 3 t゛→3. 〈(4 odd

Claims (1)

[Claims] A means for inputting the pronunciation of kanji or words, a means for storing the pronunciation input by the input means, and a kana-kanji conversion means for converting the pronunciation input by the input means into a kanji-kana mixed sentence or a single kanji. , a dictionary for supporting the kana-kanji conversion means, a means for storing the kana-kanji conversion results, and words for kanji that do not exist in the dictionary from the contents of the reading input storage means and a set of converted kanji. An automatic word learning device comprising: a word learning means for automatically learning words; and a means for automatically registering words obtained by the word learning means in the dictionary.