Visualizing Topic Flow in Students' Essays

Visualizing how the parts of a document relate to each other and producing automatically generated quality measures that people can understand are means that writers can use to improve the quality of their compositions. This paper presents a novel document visualization technique and a measure of quality based on the average semantic distance between parts of a document. We show how the visualization helps tutors mark essays more efficiently and reliably, and how a distance index calculated for the visualizations correlates with grades. The technique is further evaluated using three dimensionality reduction techniques. The results provide evidence that the degree of topic flow between consecutive sentences and paragraphs is related to essay quality.