Fujitsu Laboratories and Japan’s National Institute of Informatics (NII) will take part in an artificial brain project, known as “Can a Robot Pass the University of Tokyo (Todai) Entrance Exam?”, otherwise known as “Todai Robot.” Based on formula manipulation and computer algebra technology, Fujitsu Laboratories will participate as the math team for the project.
Led by NII professor Noriko Arai, Todai Robot was started in 2011. The goal is to enable an artificial brain to score high marks in the Japanese National Center for University Entrance Examinations by 2016, and to cross the admission threshold for Todai by 2021.
For many years, Fujitsu Laboratories has been researching formula manipulation and computer algebra for solving mathematical problems related to mathematical analysis and optimisation technologies. Fujitsu’s involvement in the Todai Robot math team is a way to develop technologies required for human-centric ICT. Technologies developed under this project should enable anyone to easily use sophisticated mathematical analysis tools, leading to solutions for a wide range of real-world problems, and the automation of mathematical analysis and optimization.
The project aims to increase the accuracy of elemental AI technologies developed to date, and integrate them to create future value in information technology, while also deepening our comprehensive understanding of human thought.
For a computer to solve math entrance-exam problems (1), it must first convert the problem text (expressed in natural language and formulas easily understood by humans), into a program executeable form. The next step is for a program known as a “solver” to solve the problem. This requires three processes as shown in the diagram below.
Getting a computer to understand text that was intended for humans is not an easy task. Using natural-language processing to draw out a semantic representation of the problem text is not simply a matter of analyzing the words. It also involves skillfully integrating mathematical terminology and a high-school level understanding of math. Beyond that, it needs to decide the best way for the computer to solve the problem. Currently only approximately 50–60% of Level 2 entrance-exam problems can be solved, even using computer algebra technology, so the challenge of the project is to improve the algorithm.
- Math Problems. Math entrance-exam problems are high-school math problems been contributed by the Benesse Corporation, Tokyo Shoseki Co., Ltd., and JC Educational Institute, Inc
- The initiatives of the math team are scheduled to be detailed in “Uniting Natural Language Processing and Computer Algebra to Solve Mathematics Problems” (Akiko Aizawa, Takuya Matsuzaki, Hirokazu Anai), a paper in a special issue on the Todai Robot project in vol. 27, no. 5 of the Journal of the Japanese Society for Artificial Intelligence.