Research

Artificial Brain Project―Can a robot get into the University of Tokyo?

Project director : Noriko Arai

This project was launched mainly by the National Institute of Informatics to provide a dream for young people. It aims at opening up a new frontier by reintegrating the artificial intelligence field, which hasbeen subdivided since 1980. Specific benchmarks are achieving a high score on the National Center Test for University Admissions by 2016 and passing the entrance examination for The University of Tokyo by 2021. We have been working to achieve these goals through this project.
In 2012 and 2013 we conducted evaluation tasks at the NTCIR, CLEF international conference. Many NLP groups participated in these tasks. In 2013, the robot took the practice examination of the National Center Test for University Admissions held by a major preparatory school for university entrance examinations. The results indicated that the robot has an 80% probability of passing exams for 400 of about 800 universities.

Cyber-physical Integrated IT Infrastructure Project to Optimize Social Systems and Services

Research representative : Jun Adachi

We have been working with universities and industry to conduct research of a cyber-physical system (CPS) at a societal scale. CPS is expected to increase efficiency in social systems and services and create new value by integrating the information system (cyber) with the physical system (physical) that functions in the real world.
The information system is used to link data from the real world that are obtained through diverse sensors with a range of information, and analyzes them. We also strive to contribute to society specifically through analysis of big data obtained in the real world, and research of a management system.

ERATO Kawarabayashi Large Graph Project

Research Director : Ken-ichi Kawarabayashi

The Internet web structure and large networks, such as Facebook, Twitter and other social networks, are expanding daily, and it is expected to reach nearly 100 billion in the near future. With this, the information volume has been increasing much faster than the advance of hardware. Promptly dealing with problems that will arise in the future presents an urgent task.
This project aims at developing high-speed algorithms that deal with probable problems in reality by taking advantage of the latest mathematical theories in theoretical computer science and discrete mathematics.