Towards Natural Language Empowered Interactive Data Analysis

Abstract

The recent advances in natural language based interaction methodologies offer promising avenues to enhance the interactive processes within the human-machine dialogue of visual analytics. We envisage Multimodal Data Analytics as a novel approach for conducting data analysis that builds on the strengths of visual analytics and natural language as an expressive interaction channel. We investigate the potential enhancements from such a multimodal approach and discusses the preliminary outline for a structured methodology to study the role of natural language in data analytics. Our approach builds on a simple model of human machine dialogue for interactive data analysis which we then propose to instantiate as visual analytics workflows - representations to study and operationalise interactive data analysis routines empowered by natural language interaction.