Aligning Divergent Contexts to Support Deictic References in Customizable Collaborative Visualization Interfaces

This is a software interface design evaluation study. Customizable software interfaces are common in data visualization. When two collaborators work together to analyze data, they may customize their own interfaces in different ways. The result is two different configurations that can be a barrier to understanding communicated deictic references (when one person wants to draw their collaborator’s attention to a specific item by pointing and saying a phrase like “look at this”). This study will have participants work in pairs to complete common data analysis tasks by using a prototype customizable data visualization system. This system is an implementation and demonstration of our proposed generic technique for measuring and reducing the divergence in context caused by differing interface configurations. When one participant wants to draw the other’s attention to a specific item, the participant can gesture and speak naturally. The system will recognize the first user’s intent, and adapt the second user’s interface to show the item the first user is referring to. The goal of this study is to collect feedback from participants on two different implementations of this interface. The results of this study will show the viability of such intent recognition and interface adaptation to assist with communicating references, and will help inform the design of future customizable collaborative data visualization system interfaces.