Comparative Visualization of User Flows in Voice Portals

Voice portals are widely used to guide users interactively through an application. Recent portals provide a growing number of functions in one application, thus increasing their complexity. This work presents flow-map-based techniques for the comparative visualization of user flows at different time frames, in order to enable dialog designers to analyze and improve the user interaction with these systems. Natural Language Systems in Voice Portals: More sophisticated voice portals use natural language systems (NLS), giving users the option to actually “talk” to the system in whole sentences. The system tries to interpret these sentences and interactively asks the user for detailed information, if necessary. Portals using NLS are rather large and complex, making it difficult to analyze their performance. Especially after applying changes to a voice portal or in case of technical problems, it is important to be able to analyze the consequences on user flows in the system.