The Allen Institute for Artificial Intelligence and Kaggle Launch a Competition to Automatically Answer 8th Grade Science Questions
SEATTLE, Oct. 7, 2015 /PRNewswire/ -- The Allen Institute for Artificial Intelligence (AI2), in partnership with Kaggle announced today the launch of a competition aimed at building a system that can automatically answer 8th grade general science questions. This competition enables AI students and researchers to try their hand at answering a large collection of multiple-choice questions using automated systems built over the course of the competition. Prizes will be awarded to the three teams who achieve the highest test scores, with a top prize of $50,000. Second place will be awarded $20,000, and $10,000 will be awarded to the third place team.
The competition starts October 7th, 2015, and will run for five months, closing in February of 2016. Winners will be announced during the AAAI 2016 Conference in Phoenix, AZ. The contest is titled "The Allen AI Science Challenge" and can be found at the following link: https://www.kaggle.com/c/the-allen-ai-science-challenge
Along with academics, AI2 encourages preeminent AI research teams in the private sector to test their systems against this challenge.
"IBM has announced that Watson is "going to college" and "diagnosing patients," said Oren Etzioni, CEO of AI2. "But before college and medical school — let's make sure Watson can ace the 8th grade science test. We challenge Watson, and all other interested parties — take the Allen AI Science Challenge."
"Of course, the same challenge holds for research groups at Google, Facebook, and elsewhere reporting cutting-edge results on deep neural networks and Natural Language Processing (NLP). I am excited to see how they will they do on the test," added Etzioni.
The Competition Design: Why Science Questions?
Project Aristo, one of the flagship projects of AI2, is actively working on understanding and answering science questions. Tackling this basic understanding of science requires building a knowledge base that also includes not just facts about science, but also elements of the unstated, common sense knowledge that humans generate over their lives. This effort represents a core building block toward mastering the fundamental AI concepts that will allow the Aristo system to evolve into a system capable of understanding and explaining science at the research level. Such a system will become an indispensable tool for scientists and researchers, helping to make new connections and discoveries as we move into an age where information is becoming so plentiful that understanding and applying that information is increasingly challenging.
"Unlike the Turing Test, standardized tests such as the one in the Kaggle competition provide us with a way to measure a machine's current ability to reason about basic scientific concepts, understand questions, and to compare its abilities with that of a human," said Peter Clark, leader of the Aristo Project. "The Allen AI Science Challenge is a way for AI2 to invite the wider community to dig into this problem with us and explore different ways to advance AI toward important, meaningful goals."
The Aristo project is described here: http://allenai.org/aristo.html.
A demonstration of some of Aristo's problem question answering techniques can be found here: http://aristo-demo.allenai.org
As part of AI2's commitment to sharing its research for the common good, all data sets and software are available for other researchers to use. See www.allenai.org/data.html.
AI2 was founded in 2014 with the singular focus of conducting high-impact research and engineering in the field of artificial intelligence, all for the common good. AI2 is the creation of Paul Allen, Microsoft cofounder, and is led by Dr. Oren Etzioni, a renowned researcher in the field of AI. AI2 employs more than 35 top-notch researchers and engineers, attracting individuals of varied interests and backgrounds from across the globe. AI2 prides itself on the diversity and collaboration of this team, and takes a results-oriented approach to complex challenges in AI.
SOURCE Allen Institute for Artificial Intelligence