Visual exploration of large collections in digital libraries
Nabani N. Silva, J. Alfredo Sánchez, Carlos Proal, Christian Rebollar · 2003
Though conceptually very powerful, Visual Information Seeking (VIS) has been demonstrated with only relatively small collections of a few thousand items, which makes it possible to keep entire collections in main memory and perform recalculations and rendering in real time as the user manipulates sliders and filters in an interactive fashion. For vast digital libraries, collections may include hundreds of thousands, millions or even more items. We describe EVA2D, a visualization environment that implements and extends the notion of Visual Information Seeking (VIS) to facilitate the exploration of large collections comprised by digital libraries. This is accomplished by introducing four main features into EVA2D: pre-computation of graphical data, simultaneous bi-dimensional selection (direct zooming), precision filtering mechanisms, and quasi- immediate feedback.