Applying Artificial Intelligence to Student Success

Concurrent Session 7
Streamed Session

Watch This Session

Brief Abstract

This session will demonstrate how artificial intelligence can be used to help improve student success through learning pathway analysis and personality dimensionalization.  Case studies demonstrating the use of AI and potential applications within the institutional context and implications will be the focus of the presentation.

Presenters

Lead Presenter: Phil Ice, Re3Tech

Phil has been in involved in online learning for over two decades. He has worked as faculty and as an administrator for higher ed institutions. He has also served as the Chief Learning Officer for Mirum worldwide, before co-founding Re3Tech. His work has been recognized nationally and internationally through the receipt of numerous awards from organizations such as OLC, DLA, USDLA, eLSE, Adobe, IBM, and others. He has over 50 peer reviewed articles and 300 presentations.

Co-presenter: Miguel Molina Cosculluela, Analytikus

Passionate about Data and what it can do for organizations today. In 2015, Miguel founded Analytikus, a Data Science & Machine Learning solutions company, aimed at simplifying how institutions consume data and solve Big Data Problems, trough predictive, prescriptive and cognitive technologies. Analytikus' solutions solve through the use of advanced analytics, the most common challenges faced by universities: Retention, Student Commitment, Recruiting and Career Planning. Before founding Analytikus, Miguel moved to Switzerland, and took the challenge of becoming World Climate Credit’s CEO. Miguel directed a digital loyalty solution (MySollars) for companies to meaningfully engage with consumers. He was chosen by Climate-KIC, a EU organization to represent Europe as a prominent Climate entrepreneur in USA. The tour included exposure to top Universities and private investors (i.e. Berkeley University, MIT, Harvard and ARPA-E among many others.) Miguel was also selected as a Climate Reality Leader, trained by USA Vice-president Al Gore and handpicked among 500 startups to present at LeWeb in Paris, one of the most important European Internet events.

Co-presenter: Melissa Layne, American Public University System

Melissa Layne, Ed.D., Associate Vice President of Research and Innovation, American Public University System. Layne earned her doctoral degree in Digital Literacies from Sam Houston State University in Huntsville, Texas and also holds a masters in Curriculum and Instruction from the University of Missouri in Columbia, Missouri. Layne's research agenda includes topics on student retention, adaptive and personalized learning, Multi-User Virtual Environments, Blockchain, Artificial Intelligence, self-paced instructional design and implementation, text analytics, informal learning, and quality assurance in online learning at the institutional, program and course levels. Her research has been recognized by several distance learning organizations including the National University Technology Network (NUTN), and the Distance Learning Administration (DLA) organization. For several years, Layne has also served on the Editorial Board for the Internet and Higher Education journal, Internet Learning Journal, the advisory council for the New Media Consortium's annual Horizon Report, and was recently appointed Editor-in-Chief of the first-ever journal devoted to OER, the International Journal of Open Educational Resources. Her work has been covered in well over 60 peer-reviewed journal publications, ten book chapters, presentations and invited keynotes.

Extended Abstract

To date, most learning analytics models and services are predicated upon student course history, demographics, aspirations (i.e. Job Role and Industry), and a potential set of courses that the student will need to complete to graduate. Though the methodologies for analysis can vary significantly, we still see a considerable amount of variance not accounted for in predicting student retention at both the course and program level. This session examines how the incorporation of cognitive-affective personality analysis, using artificial intelligence, can be used to improve model accuracy through the analysis of student work products to expose multiple dimensions of personality. This dimensional analysis is then combined with traditional data points such as those previously discussed to boost predictive accuracy.

 

As compelling as the results from utilization of this methodology are, the results are even greater when paired with other AI tools that help optimize completion planning and pathway execution. In this case, the presentation will segue into a discussion of how the LinkedIn SDK was utilized to mine professions and self-identified skills for each profession. These were then correlated with course materials for all courses at an extremely large university (over 100,000 students) using AI to conduct semantic analysis and correlation between skills and course content. The product of this analysis was individualized course maps for each student that would allow them to understand what skills were needed in their selected field. From this point the results of the personality dimensionalization analysis was used to assess where each student would likely encounter problems, allowing the institution to provide preemptive services to support students at these junctures.

 

Aside from the large institution implementation discussed above, the presentation will utilize case studies from nine other universities, where these techniques were used to improve persistence, guide career choices, strengthen student commitment, and aid in the advising process. Directions for future development and ecosystem enhancement, including the provisioning of Blockchain to provide immutable, verified records associated with the process, will be discussed. The presenters will encourage the audience to pose questions related to the implications of personality analysis as it relates to contemporary notions of self-determination in academia and the weight that this and similar tools should be given in the advising process.