Blog Post

Some questions about Voyant...

Some questions about Voyant...

After learning about Voyant through Temple Digital Scholars I decided to use it to obtain a wordcloud view of the most frequently occuring words in some of my interview data. However, this effort backfired as conversational language includes a lot of "um", "yeah", and "ya know." Is there a better way to represent interview data visually? Can you eliminate these fillers when you analyze text using Voyant? To gain some insight about the program I am studying I took the text on the homepage of their website (deleted any identifiers) and produced a cirrcus. The most frequent words are: classroom, deed, need, read, and students. This is a more academic approach to branding their organization than I expected and something to consider as I continue my research


1 comment

The short answer is to add these filler words to Voyant's standard stopword list and then tell it to use the list. there's a lot to consider in creating an appropriate stopword list for any particular project, but they make it possible to see the patterns that matter to your research by filtering out unneeded noise.