Marian Dörk, Sheelagh Carpendale, and Carey Williamson from the Department of Computer Science, University of Calgary are running a study “Visual Information Exploration on the Web” about the effectiveness of information visualization in search results. Participants of the study are asked to use a visual search prototype for a time period of their choosing and fill out short questionnaires before and after using the system. The purpose of this research is to better understand performance and usability issues of information visualization supporting search on the Web.
The visual search system you are about to use allows you to explore blog articles with the help of visualizations for time, location, and tags. The articles are aggregated from Global Voices, an online community of bloggers from around the world tracking and translating blog discussions particularly about non-western regions. Based on the blog posts’ contents, geographic locations are estimated by the GeoNames web service.
Marian describes the work behind the study as follows:
In general this research is about the potential of visualization for Web-based information seeking. In the long term we want to explore how visual overviews can benefit the search process and what kind of overviews make sense. With this particular study we want to let people use an early prototype of a visual search system which allows us to learn more about performance and usability issues. For the performance part, we’re logging delay times, bandwidth use, and query types, and with the feedback form we seek some initial reactions from information seekers about such a visual approach to search and browsing.
Participate and help establish understanding
There are no specific tasks we ask you to do as part of this study. You can freely explore the interface and go through the information. For example, you could zoom into a particular region of the world you are interested in, say Africa, select the last couple of days, and go through the tag cloud to see what kind of topics are currently discussed. Similarly, you can filter by tags and see when and where certain topics have been more active. You can combine visual filters however you want – you can also add a text query.