Page 22 - Campus Technology, January/February 2019

VIRTUAL ROUNDTABLE
5) Artificial Intelligence and
Machine Learning
Rowe: If AI and machine learning are going to impact educa- tion, the impact has to come from fundamentally redesigning every step of the educational process, from recruiting, to admis- sions, instruction, degree plans, and finally goal completion or graduation. It really is about thinking differently, and releasing the need to do all the thinking by individual or committee. We need to be open to re-imagining every process, but also em- brace what AI and machine learning can share or expose.
Lueckeman: Theresa nailed it with the need to mitigate repetitive tasks in favor of more important work — and that AI gives us that opportunity. Since the foundational require- ment to use AI is data, universities have an advantage over most companies, in general. Schools already store huge amounts of constituent data, albeit the data is siloed and of- ten spread across systems. Schools with a data warehouse that stores more than just student information system (SIS) data are in a good position for 2019.
Those that do not have a data warehouse may start us- ing their customer relationship management system (CRM) to aggregate this data — some- times even more easily than using the data warehouse — and have a ready-made sys-
tem to put that data to work. In addition to aiding stu- dent success in the classroom, AI has a host of overarching
use cases that we will see in the coming year:
Scoring. Rating and scoring data helps makes it easier to prioritize prospects, offering suggested next steps to staff and making recommendations to donors; Classification. Identification and visual recognition can transform event check-in, geofencing and even class- room attendance (so necessary for R2T4);
Service. Answering common questions with bots and virtual assistants (natural language processing) makes life outside the classroom easier for students; Automation. Using data and logic for process automa- tion minimizes manual administrative tasks — such as updating the SIS for things like address changes, ap- peals, overrides, change of major, and more — via stu- dent interfaces, bots and NLP; and
Experience. Providing modern interfaces to constitu- ents keeps departmental silos behind the curtain, en- suring students and other constituents do not have to know the “where to go” and “what to do” so common to administrative bureaucracy.
I believe that the use of AI will continue to get easier to operationalize. Many industry-agnostic, best-in-class solu- tions are finally being adopted in higher education. Tools like Salesforce.com’s Einstein Prediction Builder, as the CRM giant says, “democratize” data science, allowing those who know the data well to use it in models that are automatically selected as best for the prediction, with machine learning running in the background to provide interventions, consume results and constantly retrain the model. That means, in my opinion, schools won’t need aggregated data sets from the industry-dominating companies that are so popular right now. Instead, schools will have easier access to predictive models customized to their own data, avoiding the inherent
“Since the foundational requirement to use AI is data, universities have an advantage over most companies, in general. Schools already store huge amounts of constituent data.”
— Kathleen Lueckeman, Maryville University
“Looking ahead, we will compete with technology but win with people,” said Walmart President and CEO Doug McMillon during a company shareholders’ meeting. Connecting people who are knowledge experts and instructors with people who want to learn is a market-differentiating position in a world pro- moting online learning. Technology has to enable those connec- tions by releasing people from mundane and repetitive tasks; AI and machine learning represent an opportunity to do so.
22
CAMPUS TECHNOLOGY | January/February 2019