Mining educational data: A focus on learning analytics

Abstract

Data mining is a process of finding anomalies, implicit patterns, and correlations within large data sets to predict outcomes, or in other words, the search for relationships and global patterns that exist, but are `hidden' among the vast amounts of data. When applied to the educational domain, data mining is a powerful tool that enables better understanding of relationships, structure, patterns, and causal pathways which provide students the cognitive strategies to think critically, make decisions, and solve problems. The talk will discuss the methodology and results of this research, present the extracted knowledge, and describe its importance in the teaching-learning space. Recent developments engineered to capture and store non-cognitive affective-domain features, such as interest and persistence will be addressed. The objective is evidence-centered design and the data mining framework acknowledges that assessments entail different levels of confidence and risk.