ABSTRACT

Computers make judgements about quality all the time – for example, automated essay-scoring systems judge students’ written work, and Google’s PageRank algorithm decides which websites are of high quality. This chapter examines what has been learnt by artificial intelligence researchers in building systems to make evaluative judgements, and applies it to improving student evaluative judgement. The chapter focuses particularly on how machine-learning approaches use exemplars to learn about quality, and how these inductive approaches can lead to bias.