Can a Robot Get into the University of Tokyo? AI Project "Todai Robot" Drastically Improved Math Score Deviation Values

Achieved a deviation value of 65.8 on the Math IIB practice test

Does the word "Todai Robot" sound familiar to you? Todai Robot is the abbreviated name of a project titled "Can a robot get into the University of Tokyo?" launched at the National Institute of Informatics (NII) in April 2011 by Professor Noriko Arai. Fujitsu joined NII's mathematics team in 2012, and currently three organizations, including Nagoya University, are leading the joint research on the project.

Todai Robot's goal is to score high marks on the exam of the National Center for University Entrance Examinations by 2016 and to reach the threshold needed for admission to the University of Tokyo by 2021. The Todai Robot has taken a practice test every year since 2013, both to measure the success of the past year's work and to identify technical issues that remain. On the University of Tokyo's entrance exam pre-tests in 2013, the Todai Robot correctly answered two out of four math questions for humanities majors and two out of six math questions for science majors, acquiring a deviation value of approximately 60 among all test-takers.

This time, the Todai Robot took the Benesse Corporation's practice exam of overall scholastic aptitude. The mathematics team worked on the mathematics IA and IIB sections. The AI program automatically solved the problems, scoring a deviation value of 64.0 (75 points) on the mathematics IA section, and a deviation value of 65.8 (77 points) on the mathematics IIB section. Compared to its results in 2014, this is a drastic improvement in the deviation values of 17.1 points on IA, and 13.9 points on IIB.

Trends in test scores on Center Test practice exam (mathematics)

Scored higher with a new computer algebra program

The system the Todai Robot uses solves a problem by first converting problem texts in natural language and mathematical expressions into a form that can be executed by a program and solving the problem using a computer algebra program (a solver). Up until last fiscal year, problems were solved using a technique based on iterative transformations of expressions called "quantifier elimination" (Note 1). But this method has limitations because there are problems that cannot be calculated within the test's time limits, and areas of mathematics that cannot be handled at all. Therefore, a new approach was needed in order to get a higher score.

The improvement in the deviation values was achieved by speeding up the expression-transforming solver and developing a new solver that can handle areas of mathematics that had been off-limits.

(Note 1) Quantifier Elimination, or QE, is a technique for simplifying mathematical formulas that derives solutions by transforming problems into equivalent expressions.

The procedure of solving mathematics problems

Todai Robot is driving the development of language processing and mathematical technologies

The mathematics team on the Todai Robot project is working on ways to express problem texts in the language-processing phase, reducing the time to solve a problem, and applying automatic-solving to areas of math that were previously off limits. Work is also continuing on automating the part of language processing where human assistance had been used, in transforming problem texts to expressions.

Through the Todai Robot project, Fujitsu Laboratories will continue to work on the development of advanced language-processing technologies and mathematic techniques, and will continue its work on developing artificial intelligence technologies that take social acceptance into account for a human-centric intelligent society.

Todai Robot project math and science team (photo provided by Ashita-no-Community Lab, photographer: Seiya Kawamoto)