Rule- and case-based adaptive knowledge base and its application to Japanese-to-Braille translation
Satoshi Ono, Takashi Yamasaki, Shigeru Nakayama · 2007
Abstract: This paper proposes an adaptive knowledge base (AKB) involving two knowledge representations – rule and case. Combining rules and cases makes it possible to solve problems accurately and quickly, and to acquire new cases from problem-solving results. The proposed AKB does not require manually adjustment of the thresholds and provides higher qualified solutions than an existing method with the same knowledge source. This paper also proposes a Japanese-to-Braille translation system which uses the adaptive knowledge base as mentioned above. Experimental results have showed that the threshold adjustment reduces segmentation errors, and that the proposed system reaches almost the same translation quality as the most popular software on the market. Key–Words: Adaptive knowledge base, case-based reasoning, rule-based reasoning, Japanese-to-Braille translation 1