Abstract

Automatic Text Simplification (ATS) consists of computing techniques to rewrite text to make it simpler to read and understand, and prior research has explored its use to provide reading assistance to various user groups who may benefit from reading support. Prior literacy research has identified great diversity in literacy skills among DHH adults, and prior computing accessibility research has identified benefits from providing ATS-based reading assistance to other user groups. Thus, while ATS holds promise to benefit DHH adults, little research has investigated the benefits from, interests in and preferences for such technologies among DHH adults. Thus, this work aims to investigate the use of ATS-based reading assistance to support DHH adult readers. This dissertation is organized in three parts centered around three major contributions: requirement- and interest-gathering, methodological research, and design and evaluation work. Part I of this dissertation focuses on the requirements and interests of DHH adults on ATS-based reading assistance tools. As a relatively new technology, it is still unknown whether DHH adults would be interested in such technologies, what purposes they would be interested in them for, and what personal or technical characteristics may affect their interests. Furthermore, with various approaches to ATS available (e.g. lexical or syntactic approaches), it is unknown which ones would be of interest and/or benefit to DHH adults. Thus, we first explore perspective of DHH adults on ATS-based reading assistance tools. Then, Part II of this dissertation focuses on methodological work on how to evaluate the output of ATS technologies with DHH adults. Prior work on the evaluation of linguistic technologies has identified challenges when evaluating their technologies among DHH adults with varying literacy skill levels. However, it remains unknown if such challenges are also present when evaluating the output of ATS systems among DHH adults. Considering that there are various aspects of the output to evaluate, including its complexity or readability, and its overall quality or fluency, we investigate which metrics can reliably evaluate those characteristics of ATS output among this user group. After establishing interests and evaluation methods for various characteristics of simplified texts, Part III of this dissertation focuses on the design of these tools, and includes evaluations of such tools with DHH adults. We first investigate an exploration of the design space to identify which configurations of the various design parameters may provide more usability to DHH adults, and to understand the rational for their preferences. Using the preferred settings identified, we then conduct an evaluation to investigate the DHH users' preferences and interactions with a system that provides the various capabilities of ATS technologies using realistic output from existing systems.

Library of Congress Subject Headings

Natural language processing (Computer science); Computational linguistics; Assistive computer technology--Public opinion; Deaf--Attitudes; Hearing impaired--Attitudes

Publication Date

4-2023

Document Type

Dissertation

Student Type

Graduate

Degree Name

Computing and Information Sciences (Ph.D.)

Department, Program, or Center

Computer Science (GCCIS)

Advisor

Matt Huenerfauth

Advisor/Committee Member

Kristen Shinohara

Advisor/Committee Member

Garreth Tigwell

Campus

RIT – Main Campus

Plan Codes

COMPIS-PHD

Download

Share

COinS