Identifier

etd-06182015-152532

Degree

Doctor of Philosophy (PhD)

Department

Computer Science

Document Type

Dissertation

Abstract

Many problems related to synthesis with intelligent tutoring may be phrased as program synthesis problems using AI-style search and formal reasoning techniques. The _x000C_first two results in this dissertation focus on problem synthesis as an aspect of intelligent tutoring systems applied to STEM-based education frameworks, specifically high school geometry. Given a geometric _x000C_figure as input, our technique constructs a hypergraph representing logical deduction of facts, and then traverses the hypergraph to synthesize problems and their corresponding solutions. Using similar techniques, our third result is focused on exhaustive synthesis of molecules. This synthesis process involves bonding sets of basic, molecular `fragments' according to chemical constraints to create molecules of increasing size. For each input set of fragments, synthesis results in a significant set of molecules. Due to big data constraints we give special consideration in how to construct a corresponding molecular hypergraph based on a target, template molecule. Synthesis of the target molecule in a laboratory environment then corresponds to any path in the molecular hypergraph from the set of fragments to the target molecule.

Date

2015

Document Availability at the Time of Submission

Release the entire work immediately for access worldwide.

Committee Chair

Mukhopadhyay, Supratik

Share

COinS