Visualising Textual Concordances with Word Mosaics and Graphs: Prototypes and Evaluation Methods

The visualisation of textual concordances has seen little change since the proposal of the keyword-in-context indexing technique in the 1950s. This report details an investigation of the current state of the art in concordance visualisation. The investigation lead to the the expansion of context trees to a new graph based representation of concordance lists. This representation is explained in detail and two visualisations, of the graph, are explored. It is proposed that one of these visualisations may offer significant advantages over existing concordance visualisations for the investigation of keyword collocation frequencies. An evaluation plan to test this hypothesis is outlined as future work in this project.