CRAB Reader: A Tool for Analysis and Visualization of Argumentative Zones in Scientific Literature
Yufan Guo, Ilona Silins, Roi Reichart, Anna Korhonen · 2012
Given the rapid publication rate in many fields of science, it is important to develop technology that can help researchers locate different types of information in scientific literature. A number of approaches have been developed for automatic identification of information structure of scientific papers. Such approaches can be useful for down-stream NLP tasks (e.g. summarization) and practical research tasks (e.g. scientific literature review), and can be realistically applied across domains when they involve light supervision. However, even light supervision requires some data annotation for new tasks. We introduce the CRAB Reader – a tool for the analysis and visualization of information structure (according to the Argumentative Zoning (AZ) scheme) in scientific literature which can facilitate efficient and user-friendly expert-annotation. We investigate and demonstrate the use of our tool for this purpose and also discuss the benefits of using the same tool to support practical tasks such as scientific literature review.