This project aims to develop a tool that will support students who struggle in online learning courses facilitated by Learning Management Systems (LMSs). To that end, we will follow a two-step approach:

a) identify potentially struggling students using the log files of a Higher Education online course (data will be provided);

b) propose automatic or semi-automatic feedback interventions to support students’ learning.

In order to identify struggling students, we will engineer data-based features (Jiang, et al, 2018) that may indicate - among others - lack of background knowledge, lack of motivation, poor time-planning, and self-regulation skills. To design feedback interventions, we will use as a starting point the five levels of verbal and non-verbal interventions that tutors can employ on the learner’s progress (Wood, Wood, & Middlestone, 1978) and the concept of Contingent Tutoring (Wood, 2001).