Project Aristo

The core mission of The Allen Institute for Artificial Intelligence (AI2) is to achieve scientific breakthroughs by constructing AI systems with reasoning, learning and reading capabilities. AI2 is working to realize the vision of a "Digital Aristotle" - AI software that utilizes large amounts of knowledge in machine-understandable form to answer questions, explain those answers, and discuss those answers with users.

Overview

Project Aristo is a flagship project of AI2, a first step towards a machine that contains large amounts of knowledge in machine-computable form that can answer questions, explain those answers, and discuss those answers with users. Central to the project is machine reading semi-automated acquisition of knowledge from natural language texts. We are also integrating semi-formal methods for reasoning with knowledge, such as textual entailment and evidential reasoning, and a robust hybrid architecture that has multiple reasoning modules operating in tandem. Project Aristo represents a new start towards intelligent systems, building on experience from the prior Project Halo.

Research

The current two-year phase is giving the computer increasingly difficult science exams, at the 4th, 8th, and 12th grade levels, while aiming to gradually improve performance on, and eventually pass, these exams. This is ambitious because it requires the computer to have substantial general and simple science knowledge. Eventually, the system will evolve to field users' questions, shifting the focus from exams to those answers related to direct user queries. The three main technology thrusts within the system are:

• Knowledge Acquision: Extracting knowledge from relevant natural language texts, supplemented by other sources and techniques such as crowd-sourcing and judicious use of manual encoding.
• Machine Reasoning: Performing inference with that knowledge to answer questions.
• Question Analysis: Use of natural language processing and other techniques to interpret the actual exam questions and connect them with the knowledge.

This project is only in its preliminary stages, the first step of this endeavor is an analysis of a particular fourth grade test (the New York Regents Science Test) and its knowledge requirements, and an analysis of how well current automatic knowledge-base construction methods can meet those requirements. The system is currently existing as a partial prototype.

As shown in Figure 1, there are three main components in the system architecutre: question interpretation, reasoning, and a library of knowledge resources. Various technologies are being employed for different parts of the system, Allegro CL is being used for the question decomposition part of the question analysis phase, as well as for exploratory work for information extraction.

For additional informaiton on the Aristo Project, see here.