RMapViewer (Reaction Map Viewer) is an interactive visual environment for exploring a large network of chemical reaction paths, called the reaction map. The chemical reaction map consists of a set of atomic isomorphic 3-D structural configuration states that are theoretically derived by using the quantum mechanics. Each node is associated with its theoretically derived potential energy, and is connected with one or more other nodes via a transition link. The difference between the energy of two linked nodes represents necessary energy to transform the structural state on the one end to the adjacent state on the other end. The reaction map has been designed to support a user (i.e., a chemist) in exploring the map in a variety of ways, such as to find out possible molecule structures within a certain number of transitional steps from a focused node, or compare the energy values of possible transition paths between given two nodes. This paper describes RMapViewer by illustrating how a user interacts with the system through the different types of visualizations of the reaction map, sorting schemes, and transitional animations along with the model of a user's cognitive process in exploring the reaction map.

A zooming user interface for browsing-based archive system, KENBUN, is introduced to facilitate access the newspaper articles which are scanned from microfilm into digital images. The KENBUN provides browsing capability for the articles on the basis of the spatiotemporal knowledge of newspaper readers, where the digital images are linearly arranged according to their publish time on a two-dimensional space. We propose a zooming user interface for supporting both focused and contextual views of the space by providing the continuous and discrete interaction modes. The continuous and discrete interaction modes respectively allow the users move continuously on the space and interactively select certain times. The experiment revealed that using these modes is necessitated for browsing a large number of articles in finding not only a desired image but also text information.

The News Site Contrast (NSContrast) system analyzes multiple news sites based on the concept of contrast set mining and it can extract terms that characterize different topics of interest for specific countries. In this study, we used the NSContrast system to analyze Global Database of Events, Language and Tone (GDELT) data by comparing news articles from different regions (e.g., USA, Asia, and the Middle East). We also present examples of analyses performed using this system.