An NLP-based Reading Tool for Aiding Non-native English Readers

Proceedings of Recent Advances in Natural Language Processing, pages 41–48, Hissar, Bulgaria, 7-13 September 2013. An NLP-based Reading Tool for Aiding Non-native English Readers Mahmoud Azab Ahmed Salama Kemal Oflazer Carnegie Mellon University-Qatar Doha, Qatar {mazab,ahmedsaa,ko}@qatar.cmu.edu Hideki Shima Jun Araki Teruko Mitamura Carnegie Mellon University, Pittsburgh, PA, USA {hideki,junaraki,teruko}@cs.cmu.edu Abstract This paper describes a text-reading tool that makes extensive use of widely- available NLP tools and resources to aid non-native English speakers overcome language related hindrances while reading a text. It is a web-based tool, that can be accessed from browsers running on PCs or tablets, and provides the reader with an in- telligent e-book functionality. 1 Introduction and Motivation In this paper, we describe our approach in building a NLP-powered tool to aid in reading texts in En- glish by non-native readers of the language, espe- cially in an educational setting. Text, being bland, is hardly a conducive and motivating medium for learning, especially when the reader does not have access to aids that would enable her to get over mi- nor and not-so-minor roadblocks ranging from un- known vocabulary to unrecognized and forgotten names, hard-to-understand sentences, issues with the grammar and lack of or forgetting the prior context in a former session of reading. We aim to make reading an active and interactive experi- ence by enabling the user to interact with the text in a variety of ways using anytime-anywhere con- textually guided access to textual information. Our system is based on significant preprocess- ing and annotation of a library of texts using many publicly available NLP components for English, integrated in a UIMA (Unstructured Information Management Architecture) based server (Ferrucci and Lally, 2004). These annotated documents are then accessed via browser-based clients which es- sentially look like traditional e-book reading envi- ronments but with a much richer set of user acces- sible functionality. Thus our system can also be seen as a showcase application for demonstrating English NLP tools and resources. Our contribu- tion is the integration of many publicly available tools and resources for English into a large-scale usable application implemented in a client-server software architecture structured around UIMA, along with work on development of some annota- tion components and/or combination of available ones. In the rest of this paper, after a brief review of the use of NLP to help for reading, we will elab- orate on the user visible functionality of our sys- tem and then present the software architecture and the implementation. Our system has been imple- mented save for a couple of features and we are now in the process of planning an intrinsic evalu- ation followed by a deployment to have it be used to gauge if student users find it effective. 2 Using NLP in Reading Aids Recently, Computer Assisted Language Learning (CALL) systems have started making use of ad- vanced language technology to build intelligent systems to aid and assess reading comprehension. An early project, GLOSSER Project (Nerbonne et al., 1997) developed a system that aids read- ers …

Content truncated to 500 words