Artificial Intelligence

April 28, 2017

The New Media Institute (NMI) is a academic unit that has been growing since its creation in 2000. The program started with about 30 students but now fosters 542 students. Between the 542 students there is about 400–500 emails that get received during a semester regarding New Media classes, advising, POD, etc., which can get overwhelming at times for the professors and assistants running the program. For this reason the Artificial Intelligence team has created Xavier. Xavier is an Ai based Teaching Assistant designed to assist prospective student looking to join the NMI, and current students who need additional information. It has also been created to provide faster answers for students questions and concerns and to save professors and TA’s time since they won’t have to respond to as many emails. As the number of New Media students go up, the number of questions asked is constantly increasing, but the types of questions being asked does not really change. The same questions are generally being asked over and over again. The code that is used to create Xavier allows the AI to keep learning and growing as more questions get asked and hopefully Xavier will have a 100% response accuracy rate within a year!

Xavier is a complex program that consists of several different components. is a language recognition program that the Ai uses to analyze the data (emails) that has been received. As the data set continues to grow, helps Xavier learn more complex language and how to respond. In order for Xavier to function properly and be accessible remotely, it must be hosted on a separate server. To solve this problem, IBM blue mic cloud services is used. Finally, the optimal coding language used is node.js. These different components make the Xavier AI the helpful tool it is meant to be!

Life and projects come with many pivots, so being okay with pivoting is a huge lesson the Artificial Intelligence team has learned. Also, as time has gone on, marketers have learned that purchasing power lies completely in the hands of the consumers, and for this reason the Artificial Intelligence team had to learn to think from a user’s point of view. The team also came to realization after months of research that there is no true definition of AI. In terms of lessons learned in development the hardest lesson learned was finding the right balance between what the team wanted to do and actually being able to create it.