In an effort to make information more accessible, our team set out to refine and update a set of guidelines for clear writing and develop an initial paired text dataset to be used for improving automated text simplification. The simplification of text allows for more effective and efficient processing of textual content and the ability to automatically simplify text can make the web more accessible to everyone. Automated text simplifiers require a large dataset of paired text in order to be significantly useful. Our team partnered with IBM, UMass Boston, and UMass Medical School to create an initial dataset for automated text simplification using a refined set of operationalized guidelines for manual simplification and develop a methodology for expanding the dataset.
Worcester Polytechnic Institute
Management Information Systems
Major Qualifying Project
All authors have granted to WPI a nonexclusive royalty-free license to distribute copies of the work, subject to other agreements. Copyright is held by the author or authors, with all rights reserved, unless otherwise noted.