A Computational Approach to Edge Detection
IEEE Transactions on Pattern Analysis and Machine Intelligence
ACM SIGGRAPH 2003 Papers
ACM SIGGRAPH 2006 Papers
Proceedings of the 16th ACM SIGSPATIAL international conference on Advances in geographic information systems
IEEE Transactions on Multimedia
Hi-index | 0.00 |
We present results from two studies exploring how people use Geo-clouds, our visualization system for quickly reviewing news stories using interactive maps and extracted images from associated news images. In combination the text and image data forms a kind of collage of news information. To test this approach, we developed a Canada-centric prototype. In the first study, we assessed how familiarity with geographic region and local culture impacts the use of our system. We find that those less familiar with Canada quickly zoomed in on regions they were familiar with and remarked on image tags more frequently, while those more familiar with Canada tended to browse wider regions and to read text tags more frequently. In the second study, we explored how the ratio of text to images impacts the participant experience. We analyzed four categories of interaction behavior: zooming and panning, scanning patterns, distraction, and reading news summaries. Results show that increasing the ratio of images-to-text causes more distraction but also leads to more news summaries read. The ratio of text to images also impacted scanning behavior, with participants tending to first scan whichever tag type (images or text) was most prevalent on the map.