Zev T. Battad, RPI Graduate Student

Zev T. Battad, RPI Graduate Student

This talk will present a system for automatically generating human-friendly descriptions of numerical data graphs.  Cognitive agents are growing into roles as assistants for complex decision making and analytics, bringing large amounts of information to bear to help human analysts. Methods of narrative storytelling can benefit the ways in which cognitive agents effectively communicate such information in a human-like or human-understandable manner. In this work, we consider specifically the challenge of automatically describing numerical data graphs. We will discuss description generation focused on guiding and organizing the visual experience of data graphs, the use of narrative abnormality, and our methodology for automatic description generation.

 

Share|