Computer scientists at Fujitsu Labs have launched the "Todai Robot Project," an effort to get a robot to pass the University of Tokyo's very tough entrance exam — and they're hoping to do it in eight years.
In conjunction with Japan's National Institute of Informatics (NII), the Fujitsu researchers are striving to develop an AI that can successfully pass the infamously difficult entrance examinations at the University of Tokyo (Todai) — an institution that's considered one of Asia's best.
The project, also dubbed the "artificial brain" project, has roots going back in 2011. Recently, the AI took a sample math test from the university's entrance exams and it correctly answered two out of four math questions and two out of six science questions (some human assistance was allowed for language processing). That's a score of 40% — not bad given that it's early days.
The researchers are hoping to see the robot get "high marks" by 2016, and then get full-on acceptance into the university by 2021.
In a statement issued by Fujitsu, the company describes the technological issues that still need to be overcome:
For a computer to solve math entrance-exam problems, it must first convert the problem text, which is expressed in natural language and formulas meant to be easily understood by humans, into a form that a program can execute. The next step is for a program known as a "solver" to solve the problem. This requires three processes.
1. Language processing & semantic analysis: Understand the problem text, which is expressed as natural language and formulas easily understood by humans. More specifically, translate the problem text into a formalized semantic representation.
2. Formulation: Transform the semantic representation of the problem into a form amenable to automatic deduction. In short, convert it to a form that the computer can process.
3. Calculation: Find the answer using the mathematical solver.
Fully automating the language processing and semantic analysis part is not easy. The processing in Step 1 consists of the following three components:
1-a: Recognize the grammatical relationships between individual words (parsing).
1-b: Compose a semantic representation of the sentence based on semantic representations of the words (semantic composition).
1-c: Recognize the logical relationships between sentences(contextual analysis).
Whoa, ways to go. Though IBM's Watson, which is essentially a natural language processor and database parser, can already do much of this.