Loading project…
Loading project…
I cleaned 24,783 tweets, scored them with VADER and summed the scores. Neutral came out on top.
Original tutorial by Aman Kharwal: Twitter Sentiment Analysis (opens in a new tab)
| Output | Value | Cell # |
|---|---|---|
| Summed VADER positive | 2,880 | |
| Summed VADER negative | 7,201 | |
| Summed VADER neutral | 14,697 | |
| Overall verdict | Neutral |
Cell # counts every cell from the top of the notebook, Markdown included, starting at 1.
The dataset already carries human labels for offensive language, which the notebook ignored. Sentiment is a different question from offensiveness.
A hate-speech research dataset of tweets labelled hateful, offensive or neither.
Many tweets contain slurs and abuse. No tweet text is shown here, and the counts below are VADER sums only.