Abstract

In this study, we describe a model of Japanese honorific expressions that is constructed automatically from assembly minutes. The Japanese language has a complicated system of honorific expressions. In assemblies in Japan, speakers usually speak using honorific expression. Hence, we create a maximum entropy model for honorific expressions from assembly minutes released by Fukuoka City, Japan. Evaluations show that our model has an accuracy of 88.2% for choosing suitable honorific expressions. In particular, it shows the best precision of 95.9% for the classification of Sonkeigo at the end of a sentence.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call