Envisioning user models for adaptive visualization

Abstract

Adaptive search systems apply user models to provide better separation of relevant and non-relevant documents in a list of results. This paper presents our attempt to leverage this ability of user models in the context of visual information analysis. We developed an adaptive visualization approach for presentation and exploration of search results. We simulated a visual intelligence search/analysis scenario with log data extracted from an adaptive information foraging study and were able to verify that our method can improve the ability of traditional relevance visualization to separate relevant and irrelevant information.