Overview

Corpus Term Frequencies provides a quick look at term frequency and word distribution across a text corpus. This text analysis tool is part of the web-based Voyant text analysis workbench, developed by Geoffrey Rockwell and Stefan Sinclair. As with all the Voyant tools, you can load your own text corpus from the tool homepage using a series of URLs, or by uploading documents from your hard drive.  The tool is capable of handling plain text, HTML, and XML documents. Voyant also has a couple of pre-loaded corpora, for those who just want a quick look at the tool.

Counts and Trends

The initial screen shows all of the unique terms in the document, ranked in order from most frequently occurring to least frequently occurring. A word count is provided for each term, and total number of unique terms is displayed in the bottom right hand corner. Click on the little cog icon in the upper right hand corner to apply a stop word list that will remove common terms like “with” and “to”. A thumbnail “trends” graph appears beside each word, indicating the frequency of that word across all documents in the corpus. In a diachronic corpus (in which documents are in chronological order) peaks in the graph show spikes in usage over time.

Interface Issues

This tool is easy to use, and presents basic information about word frequency in a clear way. The addition of the thumbnail graphs is a nice touch, and the tool loads very quickly, even with large corpora. The check boxes on the left hand side are somewhat deceptive, as I was able to select only one at a time unless I used the box at the top to select all. If you wish to select a group of specific terms, use the heart icon to add each one to the favorites list, then select all favorites. When a term is clicked a new tab appears in your browser window with a much more detailed analysis of that term within the corpus. The new tab only gains focus when it is first opened, but he contents will change every time you click on a word.