Join us

We are looking for highly motivated, energetic, and independent graduate students and undergraduate students, with excellent technical and communication skills.

Current status

International Graduate Program (A) (IGP-A): Applications for this category have been closed. Please wait for the next season.

International Graduate Program (C) (IGP-C): Applications to this laboratory on this program was closed on February 28, 2019. Please wait for the next season.

MEXT Embassy Recommendation Scholarship: Applications for this category have been closed. Please wait for the next season.

General examination (for master course): Please read this page carefully especially for the program section, and refer to Japanese information (because one must pass an examination written in Japanese).

General examination (for doctoral course): Applications for this category have been closed. Please wait for the next season.

Research student: This laboratory does not accept a research student.

The status shown here does not necessarily represent the official application schedule of Tokyo Institute of Technology, but is local to this laboratory. There is a possibility that other laboratories accept applications even after application to this laboratory is closed.

How to apply

Please read this page carefully to find the best program for you and the necessary information for an application. Then, send an application to Naoaki Okazaki via email.

Fortunately, this laboratory receives quite a few inquiries from prospective students. Some inquiries cover the detailed information about a prospective student, but some include only a brief self-introduction. It is difficult for us to make a decision under the different amount of information. Thus, we kindly ask a prospective student to cover the following information in the initial inquiry.

  • Program and course (master/doctor)
  • Admission month and year
  • Summary of undergraduate research (if you did)
  • Summary of graduate research and publication list (if you are applying to the doctoral course)
  • Research plan in this laboratory
  • Result of English proficiency test
  • Transcript(s) of the past school(s)

Please use PDF format when attaching a document in an email. Do not use a Microsoft Word file (.doc or .docx) or an external link (e.g., a URL to a cloud storage); we may not open these documents for security reason.

It is highly recommended to set the sender field of your email account to your full name. We often exchange emails with multiple prospective students at the same time. We sometimes search for your emails by using your name as a query. We would miss your emails if you did not set the sender field properly.

We strongly recommend contacting multiple laboratories / universities at the same time to increase your chance. Choosing a laboratory surely change your future. Furthermore, there is rigorous competition for entry into our laboratory; we receive a number of interesting and excellent applications, but only some of which we are able to accept due to the space and resource limitation.


This laboratory focuses on Natural Language Processing (NLP), whose ultimate goal is to realize computers that understand and speak languages as humans do. Typical applications of NLP include machine translation, question answering, summarization, and knowledge acquisition. We advance the research and development of NLP on the disciplines such as linguistics, statistics, and machine learning as well as cutting-edge approaches such as deep learning. In addition, we are interested in applying the NLP technology to real world problems. Please refer to the research page for more information.

The Japanese word for research “研究” consists of 研 [ken] (sharpen) and 究 [kyu] (master). It is our hope that students in this laboratory experience the joy of research. Although they may initially wonder how a study can be a joy, we encourage each student to create their own world, complete a theory/product, and release the achievement to the public. We wish everyone in this laboratory to earn skills and confidence to be active in the industry and academia.

This laboratory provides an environment where people with various values meet; we communicate, learn basic skills, discuss the research, hang around, and occasionally talk about personal problems together. We hope that the knowledge, experiences, and relationships earned in this laboratory will be invaluable for the rest of their lives.

Expected milestones

This section explains the milestones we expect students to reach.

Bachelor study

  • Acquire the fundamental knowledge and skills
  • Read papers related to the bachelor study deeply enough to reproduce their work
  • Explain and discuss the importance and outline of the bachelor study
  • Make progress by themself through periodic meetings with their mentor(s)
  • Present the work as a scientific paper
  • Make an easy-to-follow presentation about the work
  • Present the work at a domestic conference

One might think these milestones are too much, but we are sure that they will gain a lot from a bachelor study.

Master course

In addition to the milestones for the bachelor thesis,

  • Find papers related to their study by themself
  • Propose a reasonable and new method for approaching their research topic
  • Present the work of the bachelor/master study at an international conference

Doctral course

In addition to the milestones for the master thesis,

  • Be an independent researcher who can find interesting research topics on the basis of the prior and recent studies of the research field.
  • Be a leading researcher about the research topic.
  • Present the studies at international conferences and journals multiple times.
  • Give advices for a topic, approach, paper writing, and presentation of the research of others.

We recommend reading an excellent guide, “How to Be a Successful PhD Student (in Computer Science (in NLP/ML)) by Mark Dredze and Hanna M. Wallach”.


The most convenient way for overseas students to enter this laboratory is to apply one of these programs:

Fortunately, this laboratory receives quite a few attentions not only from Japanese students but also from overseas students. However, we can accept a few students per year because of the space and resource limitation. For this reason, we make decisions about acceptance, carefully evaluating applications.

Alternatively, it is possible to apply for the standard entrance exam for the master/doctoral course. However, one must pass an examination written in Japanese on this route with the severe competition with Japanese students.

Research life

Coming to the laboratory

We do not have fixed working hours in the laboratory. Everyone can come to the laboratory and go home whenever they wants. Of course, weekends are off. We will arrange the schedule for the research seminar, study groups, and meetings with mentors by considering the schedule of each member.

Having said that, we highly recommend coming to the laboratory as possible; coming to the laboratory is definitely the first step to make progress.

Research environment

We will lend a notebook computer and monitor for all members in this laboratory. Members can buy books, devices, and software licenses when they are necessary for research, using the budget of this laboratory.


This laboratory uses Python as a default programming language. Although one can use other programming languages for their study, we recommend Python, which has a number of useful tools and libraries for data analysis and machine learning. Most new members were not so familiar with Python; we have a study group for NLP 100 exercises.

Computational environment

We maintain computation servers in this laboratory,

  • CPU servers:
    • Intel Xeon Gold 6130 processor (16 cores, 2.1 GHz) x 2, 96 GB main memory
    • Intel Xeon Gold 5118 processor (12 cores, 2.3 GHz) x 2, 96 GB main memory
  • GPU server: Intel Xeon Gold 6126 processor (12 cores, 2.6 GHz) x 2, 96 GB main memory, GeForce GTX 1080Ti x 4
  • File server: Intel Xeon Bronze 3106 processor (8 cores, 1.70 GHz 1.7G), 48 GB main memory, 40 TB HDD

In addition to these servers, we utilize supercomputers, TSUBAME 3.0, AI Bridging Cloud Infrastructure (ABCI), and Amazon Web Services (AWS).

Companies where past students entered

  • Google
  • Cygames
  • DeNA
  • IBM
  • NEC
  • NTT
  • Preferred Networks (PFN)
  • Yahoo Inc.

Frequently Asked Questions (FAQs)

Can I go to a part-time job?

If you are a foreign student who have just entered a master course, we do not recommend going to a part-time job. You may have a lot of things to learn in the laboratory after changing the environment, research topic, and mentors. Please build your brand and career in Japan.

Consider a part-time job when you have a good prospect about your study. We may offer a position of teaching assistant (TA) or research assistant (RA) (subject to the availability).

Can I go to an intern?

An internship may provide a good opportunity where you acquire the knowledge, skills, and confidence required to become successful in industry. We encourage you to go to an internship unless it interferes with your research.

I have a question about the laboratory.

Please send an email to Naoaki Okazaki.