When learning new vocabulary, both humans and machines acquire critical information about the meaning of an unfamiliar word through contextual information in a sentence or passage. However, not all contexts are equally helpful for learning an unfamiliar ‘target’ word. Some contexts provide a rich set of semantic clues to the target word’s meaning, while others are less supportive. We explore the task of finding educationally supportive contexts with respect to a given target word for vocabulary learning scenarios, particularly for improving student literacy skills. Because of their inherent context-based nature, attention-based deep learning methods provide an ideal starting point. We evaluate attention-based approaches for predicting the amount of educational support from contexts, ranging from a simple custom model using pre-trained embeddings with an additional attention layer, to a commercial Large Language Model (LLM). Using an existing major benchmark dataset for educational context support prediction, we found that a sophisticated but generic LLM had poor performance, while a simpler model using a custom attention-based approach achieved the best-known performance to date on this dataset.

Every college student registers for courses from a catalog of numerous offerings each term. Selecting the courses in which to enroll, and in what combinations, can dramatically impact each student's chances for academic success. Taking inspiration from the STEM Academy, we wanted to identify the characteristics of engineering students who graduate with 3.0 or above grade point average. The overall goal of the Customized Course Advising project is to determine the optimal term-by-term course selections for all engineering students based on their incoming characteristics and previous course history and performance, paying particular attention to concurrent enrollment. We found that ACT Math, SAT Math, and Advanced Placement exam can be effective measures to measure the students' academic preparation level. Also, we found that some concurrent course-enrollment patterns are highly predictive of first-term and overall academic success.

Early Warning Systems (EWSs) are being developed and used more frequently to aggregate multiple sources of data and provide timely information to stakeholders about students in need of academic support. As these systems grow more complex, there is an increasing need to incorporate relevant and real-time course-related information that could be predictors of a student's success or failure. This paper presents an investigation of how to incorporate students' use of course resources from a Learning Management System (LMS) into an existing EWS. Specifically, we focus our efforts on understanding the relationship between course resource use and a student's final course grade. Using ten semesters of LMS data from a requisite Chemistry course, we categorized course resources into four categories. We used a multinomial logistic regression model with semester fixed-effects to estimate the relationship between course resource use and the likelihood that a student receives an "A" or "B" in the course versus a "C." Results suggest that students who use Exam Preparation or Lecture resources to a greater degree than their peers are more likely to receive an "A" or "B" as a final grade. We discuss the implications of our results for the further development of this EWS and EWSs in general.