https://juliasilge.com/blog/tag-network/

Prophet Forecasting Ad Revenue.

Top 5 Exchanges/Coins

Stacked Box Plot showing resolution via Browser type

https://netpeak.net/blog/how-to-measure-lost-revenue-in-google-adwords-with-the-help-of-language-r/

By Device and Channel using Google Analytics API and using a few libraries to get a Tree with a Channel Grouping by Device report based on Sessions

1,000 tweet locations using the hashtag #puertoricorelief

Upper & Lower Bounds included using a segment for organic traffic based on historical 3 years. Always good to capture at LEAST 3x the amount of data that you are planning on forecasting. If you want to forecast 1 year, then grab the last three years and make sure it is aggregated down by day. If you do it higher level, the more variability will occur and you'll be very sorry... Advice given.

Text mining transcribed phone calls using Google Speech API and R with textmining package as well as ggplot and others- but this one in particular is using the Wordcloud package and RColorBrewer

Example

When do people tweet (and retweet) about Home Depot

You can see the one negative plot that is coordinated below the base, and the positive, although most were neutral in the analysis

What I've done now - is I've essentially weighted, ranked, both hierarchially and associatedly the words into specific categories and modeled them via "Topic Modeling" method.

Cluster dendogram of 6 clusters. It is grouping words by weight, association and hierarchically

This was interesting. Now only if you could zoom in. I can though...

Word Frequency Ranking, basically...

If you saw my post. Then you know what this wordcloud is of.

Using Google Maps API and Twitter API

Topic modeling can be used to categorize words into groups. Here, we can use it to see whether (some) of my followers can be grouped into subgroups according to their descriptions.

What are the most postive and negative words via the sentiment analysis in my twitter follower's descriptions. Demonstrated here.

Most of my followers have predominant positive sentiment in their descriptions- this is important for the following graph to see how it all comes together.

Basically most common "paired" descriptions which in this case having a lower sample, this is not a surprise looking at it.

Need I say more.

What are the most commonly used words in my followers’ descriptions?

Yes. But maybe if I had more followers...

I also have information about the number of followers that each of my followers have (2nd degree followers). This shows that my followers have a descent amount of followers compared to other profiles

Need I say more?