Computational Science Initiative Event

A vast amount of information and data is generated every day from various sources relevant to work or to personal lives. There are several types of data such as 2D spatial and non-spatial data, 3D volumetric data, multivariate data, and crowdsourced data. Depending on the type of the input data, different data operations, encoding, and interaction techniques are needed. Visualizing such data is targeted to various users, including both visualization novices and experts. Not every user desires the same level of data analysis or has the same background knowledge and experience in visualization. Depending on the data characteristics and the target users, several different factors such as usage context and aesthetics are considered in visualizing the data effectively. We present user-centric visualization frameworks, each of which was designed to visualize and/or process various data by understanding target users (non-experts, experts) in a different perspective.
First, visualizing personal data from various sources such as sensors has different requirements from traditional visualization applications. One of these requirements in personal visualization is aesthetics. We present two frameworks (Ambienizer and Graphoto) using photos to help users understand their personal data with aesthetically pleasing displays. In Ambienzier, information is mapped through image processing techniques, which are applied to a photo. In Graphoto, we generate a photo or adjust an existing one to match a line graph as a visual embellishment of the graph.
We also introduce a visual framework to explore multivariate spatio-temporal data for domain experts. In the framework, the input data is encoded into glyphs and several interaction techniques are provided to help experts obtain an overview of all regions and insight into selected local regions. Compressing and visualizing volumetric data generated by certain devices such as CT and MRI scanners on mobile devices is still challenging bec