Standardizing Complex Functional Expressions in Japanese Predicates: Applying Theoretically-Based Paraphrasing Rules
Tomoko Izumi, Kenji Imamura, Genichiro Kikui, Satoshi Sato · 2010
In order to accomplish the deep semantic understanding of a language, it is essential to analyze the meaning of predicate phrases, a content word plus functional expressions. In agglutinating languages such as Japanese, however, sentential predicates are multi-morpheme expressions and all the functional expressions including those unnecessary to the meaning of the predicate are merged into one phrase. This triggers an increase in surface forms, which is problematic for NLP systems. We solve this by introducing simplified surface forms of predicates that retain only the crucial meaning of the functional expressions. We construct paraphrasing rules based on syntactic and semantic theories in linguistics. The results of experiments show that our system achieves the high accuracy of 77 % while reducing the differences in surface forms by 44%, which is quite close to the performance of manually simplified predicates. 1