These studies have been accomplished at Department of Information Science, Kyoto University, Japan, since I entered the doctoral course.
I would like to express sincere gratitude and appreciation to Professor Shuji Doshita for his guidance and encouragement during the last four years. I would also like to cordially thank Associate Professor Tatsuya Kawahara for his unflagging efforts to compete with world-class international research, which made this thesis possible. I am also very much obliged to Professor Masatake Dantsuji at Center for Information and Multimedia Studies for his valuable advice and comments on linguistics. Special thanks to Dr. Masahiro Araki for answering my numerous questions. I gratefully acknowledge non-native learners of Japanese and native speakers for their cooperation in the experiments.
Many colleagues have influenced this thesis, including past and present members of the Doshita laboratory and the Kansai Speech Group having seminars biannually. I especially want to thank Mr. Akinobu Lee for the maintenance of computer systems. Additional thanks to Dr. San-Sung Seomun at Tokyo Institute of Technology and Mr. Chang-Kyun Han at Tokyo University - who came together to Japan in 1995 as Japanese government scholarship students - for their contributions and their friendship.
I especially wish to thank Professor Chang-Hoon Lee, who was my academic advisor in the master course at Department of Computer Engineering, Konkuk University, Seoul, Korea for his concern and kindness. Special thanks to Mr. Sang-Bong Yang, Mr. Seung-Hoon Ra, Mr. Jung-Whan Ji, Mr. Byung-Man Park for their friendship. Additional thanks to Mr. John Brewer, my English teacher and best friend, for his support. Finally, I wish to thank my parents, my elder brother and younger syster for their constant love.
