P05 - The Chatbot Update System (CUS): An Effective Interface to Train AI
Description
With the rise of AI in many disciplines and the proliferation of chatbots in many applications, various chatbots need training to properly respond to human users. In this presentation, I report on a chatbot training interface that I developed named CUS, the Chatbot Update System. CUS was developed for use with a cybersecurity playable case study that immerses users in an experience like unto working in a cybersecurity firm. A chatbot plays the users’ coworkers in the simulation, and the chatbot needs training to recognize the meaning of various user inputs. CUS successfully provided a convenient and efficient way to provide appropriate responses to user input. With this presentation, I show the most recent version of CUS, which includes new features: gamified elements, small sets of corrections, a mobile-friendly interface, and an arbitration feature.
Presenter(s)

Presenter
Stephen Francis is a PhD student at Brigham Young University. He is a programmer by trade and an instructional designer and psychologist by education. He loves interdisciplinary work and has created an app to improve AI, a game in VR, and web apps to perform various tasks. He is always looking to improve efficiency, effectiveness, and quality.