2015, or surprisingly, a decade ago
In 2015, the most popular data projects had to have been Twitter sentiment analyses. People LOVED trying to use tweets to predict stock prices or US presidential elections.
And if you met me a decade ago, you might know me more as a data guy than a software guy (though I realize those lines are heavily blurred these days). In that previous life, the majority of my days were spent in Excel, Access, or R/SAS. Professionally, I've since gone a different direction but sometimes I backslide a little...
2025, back to the future
So I'd like to celebrate a decade of programming by marrying my two disciplines (coding and data analysis) and dive into a good ole fashioned Twitter dataset. The dataset we'll play with is New Years Resolution tweets from 2015.
This time around, I can make the engineering a bit more interesting so I built out a command line interfaced backed with a simple search index and a character by character trie implementation.
That's right folks - it's a Twitter analysis with search autocomplete!
Here's a few gems...Run it yourself here: github.com/stoneG/twitter-resolutions