Evaluating the Benefits of Explicit and Semi-Automated Clusters for Immersive Sensemaking
Ibrahim Asadullah Tahmid, Lee Lisle, Kylie Davidson, Chris North, Doug A. Bowman
Abstract
Immersive spaces have great potential to support analysts in complex sensemaking tasks, but the use of only manual interactions for organizing data elements can become tedious. We analyzed the user interactions to support cluster formation in an immersive sensemaking system, and we designed a semi-automated cluster creation technique that determines the user’s intent to create a cluster based on object proximity. We present the results of a user study comparing this proximity-based technique with a manual clustering technique and a baseline immersive workspace with no explicit clustering support. We found that semi-automated clustering was faster and preferred, while manual clustering gave greater control to users. These results provide support for the approach of adding intelligent semantic interactions to aid the users of immersive analytics systems.