Big Data and Artificial Intelligence in Higher Education
Dr. Russell Pinizzotto
Former Interim Provost and Vice President for Academic Affairs
Carlow University (Pittsburgh, Pennsyvania)
According to the software development company ScienceSoft, education lags behind most sectors of the economy in the use of big data, data analytics and artificial intelligence (AI). And yet, these technologies can significantly enhance the effectiveness of educational institutions. Big data helps educators at the University of Central Florida understand trends in student success, and the University of Alabama uses it to identify students at risk of leaving. At Dartmouth College, big data helps inform instructional design. Georgia Tech, Staffordshire University in the UK, and the University of Murcia in Spain all use AI-enabled 24/7 chatbots to answer students’ campus and curriculum-related questions. urcia researchers recently found the chatbots answered correctly more than 91% of the time and provided ancillary benefits in increased student motivation and improved staff productivity.
Applications in Higher Education
Academics, student affairs, and institutional operations are all fertile ground for these methodologies. For example, enrollment management functions typically have significant datasets that can be used by AI systems to analyze pre-enrollment data. These analyses can then be used to inform admissions decisions or to set scholarship amounts that result in significant institutional savings with no erosion of enrollment rates. Robust AI models are capable of predicting retention and graduation, especially when combined with post-enrollment data such as grades and academic progress. It is even possible to predict prospective students’ college grade point averages (within limits) with a surprisingly high degree of reliability. AI can be used to monitor students’ academic progress, adjust personalized instruction, and monitor engagement, which is a good predictor of retention. Armed with these sophisticated tools, institutional resources can be allocated, and even targeted in real time, more appropriately.
Big data analytics and artificial intelligence lie at the intersection of computer science, information technology, advanced statistics, and institutional operations. However, the basics are relatively straightforward.
Big data is characterized by the “three Vs,” volume, variety, and velocity. The data is so large, so complex and/or occurs so rapidly that it is hard to deal with using standard methods. The data may be structured or unstructured, and occur not just as numbers, but also in text and video. Big data analytics make it possible to gather data from multiple sources in multiple formats and then process it to discern the information that lies within.
Big data analytics use highly automated methods of examining data to obtain information and reach conclusions. Analytics are generally split into four categories of increasing complexity and utility to institutional decision making. Descriptive analytics focus on past events and answer basic questions about what happened and when it happened. Diagnostic analytics try to address why events occurred. Predictive analytics forecast what is likely to occur in the future. Prescriptive analytics suggest actions to take to optimize a process. Data analytics are not only used to discover and interpret information, but also to communicate results in easily understandable and visually compelling ways to institutional stakeholders and decision makers.
Artificial or machine intelligence refers to computer systems that mimic the cognitive functions usually associated with human intelligence, such as visual perception, speech recognition, learning, and problem solving. Perhaps the best-known example of AI is IBM’s Watson, the AI that famously beat Jeopardy’s Ken Jennings and Brad Rutter to win a $1 million championship challenge. AI systems improve over time by using examples to learn. AI commonly uses automated model building without specific programming. AI is not as futuristic as it sounds: your last interaction with a customer service “representative” during a live chat was most likely an AI system.
Big data, data analytics, and artificial intelligence are powerful tools that must be embraced by higher education to help improve productivity and institutional decision making. Institutions that learn to leverage these methodologies will have a distinct advantage in higher education’s increasingly competitive marketplace.