Visual time period analysis: a multimedia analytics application for summarizing and analyzing eye-tracking experiments

摘要

Recently, an increasing need for sophisticated multimedia analytics tools has been observed, which is triggered by a rapid growth of multimedia collections and by an increasing number of scientific fields embedding images in their studies. Although temporal data is ubiquitous and crucial in many applications, such tools typically do not support the analysis of data along the temporal dimension, especially for time periods. An appropriate visualization and comparison of period data associated with multimedia collections would help users to infer new information from such collections. In this paper, we present a novel multimedia analytics application for summarizing and analyzing temporal data from eye-tracking experiments. The application combines three different visual approaches: TIME. DIFF, visual-information-seeking mantra, and multi-viewpoint. A qualitative evaluation with domain experts confirmed that our application helps decision makers to summarize and analyze multimedia collections containing period data.