Date of Award

May 2023

Degree Type

Thesis

Degree Name

Master of Science

Department

Computer Science

First Advisor

Susan W. McRoy

Committee Members

Tian Zhao, Ethan V Munson

Keywords

reading comprehension, reading disabilities, recipe named entity, text simplification

Abstract

Recipes can be important for accomplishing preparatory activities for eating, a major life activity and sometimes an important social occasion. Disabilities that cause reading problems that impact comprehension can make it difficult for an individual to read and understand recipes. A simplified recipe would reduce the mental effort required to read and execute the recipe. There are resources for individuals who have more severe impairments, but these are not suitable for individuals with more mild impairments. A proof-of-concept software system for extracting information information from cookie recipes was created. The system uses a combination of rules and a statistical model to make a recipe's oven preheat temperature, baking duration and required tools easier to identify; format actions in bold typeface; and separate multi-sentence steps into sub-steps. The statistical model uses named entity recognition. The system was evaluated in a pilot user study.

Available for download on Monday, February 17, 2025

Share

COinS