| 摘要 |
PURPOSE:To deal with translation processing including spelling checkup in one stroke for shortening the processing time and saving a storage resource by adding access information of morpheme analyzing dictionary to information of each vocabulary of a spelling check dictionary. CONSTITUTION:Translation object English sentence data is stored from an English sentence input function 1 to a storage function 2, a token split function 3 extracts token and stores the extracted token in a translation data storage function 4. Next, all of word form changes are considered to find out a table using all of surface layer notations of words for heading is retrieved using a table retrieval function 6. From this retrieval result, if a relevant entry cannot be found out, it is concluded that there must be a spelling mistake in a token used for the retrieval. If a relevant entry is found, an address of the morpheme analyzing dictionary 9 where the relevant entry found was described is added as additional information to the token. Thus, time taken for translation can be shortened including spelling checkup and a storage resource can also be saved. |
