To convey shape images of Kanji characters to blind persons, we propose "kouseiyomi" or constitutional explanation of kanji characters. Kouseiyomi of a kanji character consists of names of the parts of the kanji and their positional relationships. We made a set of rules of breaking down kanji characters into parts and developed a set of kouseiyomi for 6,355 kanji characters included in JIS 1 and JIS 2 lists based on these rules. We performed an experiment in which sighted participants were asked to write kanji characters expressed by kouseiyomi. The result shows that they could write kanji at a correct rate of around 90%, which is significantly higher than the rate performed with explanation by idioms. This result demonstrates the effectiveness of kouseiyomi in conveying shape images of Kanji characters. At the same time, a few kouseiyomi expressions were answered with low correct rates. We considered its reason.%漢字の形を正確に知りたいという視覚障害者の要望を受けて，漢字を言葉で伝える「構成読み」を提案した．構成読みは，漢字を構成する複数の部品とその位置関係で漢字の形を表すものである．漢字の分解ルールを定め，これに従って，JIS第1水準・第2水準の6355字の漢字を作成した．構成読みによる漢字想起の精度を調べるため，構成読みで表した漢字を書かせる実験を晴眼者を対象に行った．その結果，熟語による説明では書けなかった漢字も約90％の正答率で書くことができた．これにより，構成読みの有効性が示された．一方で正答率の低い構成読みも一部あった．その要因について検証した．
展开▼
