Evaluating Conversational AI Systems: Constructs, Metrics and Solutions

Write your awesome label here.

Over the last few years, dialogue systems in Conversational AI have become much more robust. But to go from an ‘ok’, to a delightful experience your team still needs structured evaluation. Selecting the right approach remains a challenging task.

Over the last few years, dialogue systems in Conversational AI have become much more robust. But to go from an ‘ok’, to a delightful experience your team still needs structured evaluation.

However, selecting the right approach remains a challenging task.

In this webinar we will provide an overview of the used constructs (what can you measure?) and metrics (how can you measure this?) found in previous work, and discuss challenges (and potential solutions) in the context of dialogue system evaluation.