|
Abstract: In this talk I highlight three (out of many) trends in InfoVis Research; describe the current state of affairs, and suggest possible next steps in advancing the research agenda. The trends are:
1. Democratization of InfoVis-driven in large part by Business Intelligence and Data Analytics. Tableau and Watson Analytics are state-of-the-art commercial systems. Multi-modal and more sophisticated natural language interfaces, ever-smarter automatic generation (Show Me on steroids) and proactive user guidance are promising directions.
2. Ever-improving "programming" tools at ever-higher levels of abstraction: APIs for Java and Processing, Prefuse, d3, Vega. Could programming by example be the next step?
3. Automated creativity. Our community is continually developing novel visualizations - see ChinaVis, EuroVis, VisWeek, the New York Times, Wall Street Journal, etc etc etc. Suppose we could capture the creativity (design thinking) embodied in many visualizations - could Machine Learning and Crowd Sourcing be harnessed? Could creativity perhaps be algorithmitized? (That would put a lot of us out of work.) |