Algorithmic Generation and Mobile Distribution of Phonetic, Orthographic, and Inference-Based Literacy Exercises for Adult Learners.

摘要

A 2014 study found that more than half of all Americans aged 16-74 do not possess the literacy skills needed to adequately cope with the demands of day-to-day life, with over 13 million people in the United States alone considered to be functionally illiterate. There have been numerous efforts to address these high illiteracy rates by way of Adult Basic Education programs; however, such programs are consistently underfunded and plagued with high drop-out rates, as many students struggle to consistently attend classes due to lack of transportation, uncontrollable family circumstances, or conflicting employment hours. At the same time, smartphones are becoming ever more ubiquitous, signifying an opportunity for mobile educational software to address many of these difficulties.;This thesis describes the creation of a software system to allow students in such programs to continue their learning outside the classroom. This project, known as CAPITAL (Comprehension and Pronunciation Instructional Tools for Adult Learners), involves three core components. The first is a system for algorithmically generating questions to thoroughly cover the different microskills of alphabetic literacy, the beginner level of reading. This system is capable of generating materials for testing every facet of phonological awareness, phonemic awareness, decoding, and encoding, customized to accommodate any desired skill progression. Second is a system for algorithmically generating questions for reading comprehension at the intermediate reading level that specifically target the reader's ability to draw inferences and monitor their own comprehension. The final component is a mobile-learning application, carefully designed for ease of use by adults with below-average literacy, through which students receive these generated learning materials. Trial studies were conducted with literacy providers in the Washington D.C. area to seek feedback from both teachers and students.;The combination of these three components results in a mobile application that is easy for adult learners to use, provides them with a substantial amount of high-quality learning materials, seamlessly integrates with the lessons they are learning in their classes, and delivers materials to them at an optimal and customized rate. CAPITAL is the first system of its kind to address the adult literacy crisis by both allowing for scalable automated creation of learning materials and providing a more effective method of delivering them to students in need.