SBIR Phase I: Software application for conversion of text-heavy documents into interactive diagram-based documents

This SBIR Phase I project addresses reading and learning difficulties faced by many learners in school, college, and the workplace. The goal of the project is to build a web application that helps individuals and organizations convert text-heavy documents (such as online textbooks, manuals, and courses) into interactive diagram-based documents. The greatest potential benefit will be for students who have difficulty learning from long paragraphs of text ?including visual learners, English Language Learners, and students with dyslexia or ADHD. The users of the app will be consumers of educational content (students and online learners) and content creators (authors, publishers, test prep companies, and corporate training departments). The theoretical foundation of this proposed research comes from studies that show the learning benefits of graphic organizers in a wide range of disciplines, including science, engineering, humanities, and social studies. This proposal will support the nation's need for a skilled workforce by helping to increase the number of students who graduate from high school, college, and post-secondary vocational programs and by improving the learning outcomes for employer training and retraining programs to meet changing business needs. The commercial impact is expected to be strong, generating income for tax revenue and creating jobs. The technical innovation of the project is a methodology that provides a well-defined sequence of steps for semi-automatic conversion of complex text-based documents to interactive diagram-based documents. Fields of knowledge to be drawn upon to build the web application include instructional design, information visualization, text mining, and natural language processing. The major technical hurdles to be addressed are accurate identification and extraction of content from a source document, displaying the content using interactive visualization techniques, and integration of the converter with the Company's prototypes of a viewer and an authoring tool (built using HTML5, CSS3, and JavaScript). One way that the methodology lowers the risks involved in the conversion process is to separate the text-mining steps (relatively low risk) from the natural language processing steps (relatively high risk). If successful, the converter will dramatically reduce the time and cost for individuals and organizations to convert existing educational content into documents that are more accessible for diverse learners. The scope of this proposal is limited to the conversion of digital textbooks that are available as open educational resources. Future research will build on this methodology to convert other types of documents (e.g. manuals and online courses) and file formats.