| 摘要 |
PURPOSE:To improve conversion correctness with sure and ease by retrieving a work within the range where a semantic information conincides with that of the word in other clause, when discriminating a homonym and converting it. CONSTITUTION:In connection with RYOUYOKOZUNA (KANA, in Japanese syllabary), for example, a position data number 10 which corresponds to the prefix having a pronunciation RYOU is read out of prefixes in a storage area 21, an corresponding autonomous word number 100 is read in a storage area 22, the autonomous words attached with the autonomous word number 100 in a storage area 23 are sequentially retrieved, and then, YOKOZUNA (in Japanese syllabary) is selected and converted into RYOUYOKOZUNA (KANJI, in chinese character). In a semantic area 34 of the storage area 23, those having similar semantic information are grouped together among the autonomous words which are possible to be coupled prefixes (for example, a which means man, b which means time), and its group identifying data is registered. In connection with conversion of a KANA string that spans multiple segments, those with similar semantic relation among clauses are given priority for conversion. |
