word_cloud VS scattertext

wordcloud to make the word clouds

- Logo with WordCloud. The texts used were from the plot/synopsis from the Wikipedia pages of the top 20 anime according to MyAnimeList (excluding anime from the same series. I'm looking at you Gintama).

Here's a GitHub issue that tracks this.

Word arrangement generated using the wordcloud Python package with a custom placement mask (skull design by me) and custom color function. Common English words removed before processing.

After reading this tutorial you should now be able to generate your own Wordcloud using Python. Checkout the WordCloud for Python Documentation. Use your imagination and have fun. Please leave like or comment if you found this article interesting!

If you are proficient with Python then there is also the wordcloud library (and its homepage/documentation) to make a custom word cloud for yourself rather than relying on TagCrowd. For example, I noticed that using the HasteBin as a URL source caused TagCrowd to put some gibberish words in the cloud, like “donaeurtmt”, which do not appear when simply pasting the raw text into the text field.

The project is very heavily based on https://github.com/amueller/word_cloud

A quick and easy way to get some context into anything being mentioned ensure the CSV is only the responses and run it through a word cloud generator in Python to determine what those keywords are. That should help you see what your customers are at least mentioning and then go from there to look for surrounding context.

If what you want is to determine how similar two categories are, or to learn something about the structure or words that compose those categories, you might consider word shift graphs or Scattertext.

Other suggestions: Take a look at Scattertext. Compare keywords to the problem of aspect extraction. I think an underutilized way to look at textual data when you have a single group of interest is the word-frequency-based odds ratio.