This is an interesting and useful post. Bodong Chen first distinguishes between learning analytics and academic analytics (the former being directly concerned with teaching and learning) and educational data mining (the latter being more focused on the exploration of data from academic settings). He then outlines some areas of interest: first, the emphasis on big data in learning analytics, and second the need for it to consider the more nuanced aspects of learning itself. This leads to a discssion of the 'tensions': first, divisions based on different accounts of learning; second, the tension between learning and algorithms; third, agency and control; and fourth, the ethics of learning analytics. It's also worth viewing his Learning Analytics course.