Tesserae: A Search Engine for Allusion
The early stage development of a computational tool to detect and analyze literary allusions, with an initial focus on Latin and ancient Greek.
The Tesserae Project is an interdisciplinary research effort employing computational methods to detect and analyze literary allusion (a form of text reuse) currently focusing on Latin and ancient Greek. The Project seeks funding to create a fully-functional, publicly available tool to detect similar phrases in two texts at rates that approach those of literary commentators. To this end, funding will support adding sensitivity to word meaning, phrase context, and sound similarity. Detection rate improvements will be measured against a set of 3000 parallel phrases previously graded for literary significance. A revised website will inform researchers of research results and new functions of the tool. The project team will give presentations and produce publications explaining the function, results, and theoretical consequences of the fully operational tool. This work is preliminary to an out-year Implementation Phase that will see the addition of English, French, Italian, and Spanish.
[White paper][Grant products]