Topic visualization for understanding research paper in collaborative discussion

Abstract

Our objective is to construct a collaborative discussion support system for understanding the research paper in the research group by visualizing the contents of the discussion. When reading a research paper, it is essential not only to understand the contents of the paper but also to obtain the related knowledge. Our system evaluates topics in discussion from the viewpoints of similarity between a topic and the paper and that between topics. Then, the system visualizes topics as circles around chapter nodes whose distances from their target chapter nodes and other topic nodes represent the similarities. By observing the visualized topics, to derive topics from various perspectives is encouraged. © Springer-Verlag 2010.