Text Mining Exercise

For this exercise, I picked Wild Kindred by Jean M. Thompson at random to process - I wanted to see what I can gather about the text solely from the output of the text mining tool.

Text Analyzer

From Text Analyzer, I was given key words that are used in the text, which all seemed to do with animals and nature. I gathered from this, and inferred from the title, that the text was most likely a narrative about nature. I thought it was interesting that Text Analyzer recommends other texts based on the results of the processed text. It seemed to find connections with more informational, rather than narrative, texts about species of animals due to the key words. This seems to show the limitation of text mining using bodies of text as “bags of words” since deeper analysis is needed to uncover the genre of the text and make more relevant recommendations.

Voyant Tools

I also tried using Voyant Tools, which has a very different output than Text Analyzer. It seems like Text Analyzer has a more sophisticated word filtering algorithm that focuses on nouns, while Voyant Tools appears to have less filtering and thus considers adjectives, adverbs, and articles as well. However, Voyant Tools does a deeper investigation into factors such as where along the text words are used. It seems like Text Analyzer is better at finding the topic of the text, but combined with the results of Voyant Tools perhaps a deeper understanding of the text can be uncovered, such as sentiment or genre.