2020年第4期

【Abstract】Japanese’s quasi-characters iejirushi, literally ‘house sign’, are traditional marks or logos for companies in Japan. In most cases, they involve one or a few Japanese characters, combined with other, various, non-character symbols. For different reasons as explained in this article, quasi-characters have been ignored by character encodings such as unifying ones like Unicode or local standards like JIS X 0213, and thus pose an information representation issue for computer systems. In this paper, building on previous works, we first conduct a detailed ontological analysis of Japanese’s quasi-characters in order to identify their characteristics, especially their contents and structures. This is a major part of this work. Second, we derive from this ontological discussion possible representation methods for computer systems and conduct some experiments. The experimental results show significant improvements with respect to quality, practicability and flexibility over conventional approaches.