Predicting Academic Performance Based on Social Activities

Abstract: Predictive modelling is an important part of learning analytics, whose main objective is toestimate student success, in terms of performance, knowledge, score or grade. The data used for thepredictive model can be either state-based data (e.g., demographics, psychological traits, past performance)or event-driven data (i.e., based on student activity). The latter can be derived from students' interactionswith educational systems and resources; learning management systems are a widely analysed data source,while social media-based learning environments are scarcely explored.Data is collected from a Web ApplicationsDesign course, in which students use wiki, blog and microblogging tools, for communication andcollaboration activities in a project-based learning scenario.In addition to the novel settings and performance indicators, an innovative regression algorithm is used for grade prediction. Very good correlation coefficients are obtained and 85% of predictions are within one point of the actual grade, outperforming classic regression algorithms.

Keywords: component, formatting, style, styling, insert (key words)