2025Year

First Semester

Computational Linguistics

Lecture

Computational Linguistics

Z1100014 Computational Linguistics

KONDO Tadahisa

2.0Credits

Tue.4Period

Hachioji Remote

A) A high degree of specialized expertise　50%
B) The skills to use science and technology　50%
C) The ability to conduct research independently, knowledge pertaining to society and occupations, and sense of ethics required of engineers and researchers　0%
D) Creative skills in specific areas of specialization　0%

Computational Linguistics and Natural Language Processing (NLP) are impressive research fields of computer science. This course provides knowledge of basic processes (main modules) of the natural language processing system and knowledge of mechanism of the human language processing with several research methods in psycholinguistics.
Students acquire skills to consider deeply by themselves and discussion with other members how to
solve the remaining problems of the fields.

It is desirable, but not required, to have completed "Computational Linguistics", a third-year course at the Department of Information Design, Kogakuin University.
This class starts from survey of the course.

Additionally, students discuss and explore new directions in human language processing, language understanding, and psycholinguistic research.

Discussion Debate／Group Work／Presentation／Support for self-learning using ICT

1. Guidance (What is Natural Language)
2. Characteristics of speech, character, and text of Japanese
3. Japanese corpus and text processing (Language Resources, Big data)
4. Morphological analysis and dictionary (module 1)
5. Parser and grammar (module 2)
6. Semantic analysis and higher level processing (knowledge representation)
 *A midterm report is assigned
7. Review of the NLP technology (brainstorming and discussion)
8. Applications and advanced research of NLP
9. Human language processing (word, lexicon, reading)
10. Psycholinguistics and Corpus linguistics
11. Speech production（production plan, word order, prosody）
12. Language acquisition and learning system
13. NLP technology and human language processing (debate)
  *A term-end report is assigned
14. Review of the course contents

Students must submit two reports at the midterm and the end of term. The final evaluation would proceed by the grade of the reports (70%) and activities in the class works (30%). Refering to the GPA, D grade would be a standard mark.

Student's performance level are given as comments and suggestions for reports and discussion or debate time.

Texts are not appointed particularly.

Hagiwara et al. (transl.), 2010, “Nyumon Shizengengo-shori”, Oreilly Japan.
Francis Bond et al. (ed.), 2016, “Readings in Japanese Natural Language Processing,” CSLI publication.
Additional references are appointed in some classes.

The office hour is the 4th and 5th period on Wednesday at the lab(2-606) in Hachioji campus and the lecture room light after each class. Also, students’ visits to labs (both in Shijuku and Hachioji) with an appointment and e-mails are welcome.

The language is the most important media for human communication. Computer systems need complicated processes to understand the message of inputs by the NL because of flexible linguistic performances. Therefore, there are still many problems remaining in NLP. The mechanism of human language processing is also still mystery. Students who have a will to challenge solving the remaining problems and mysteries are welcome.

Applicable

研究開発，企画の経験がある教員が、研究調査計画，実施，発表までの経験を活かし、専門知識，考察，ディスカッションについて講義する。

Informatics Program