RE: Honouring My Steemians - Mentor @DNews
Thanks for the post, d00k13.
I hope you don't mind if I test out some sentiment analysis on your post. This is an experimental bot running on posts that have exceptional positivity or negativity. The goal is to iterate towards a bot that gives content creators and curators actionable and useful information.
Your post was selected because it is in the 99th percentile for positivity.Your post had an average negative sentiment of -0.03, an average positive sentiment of 0.206, and an average normalized sentiment of 0.18
The most positive sentence in your post had a normalized positivity score of 0.666:
"If you wish too help, Power please!"
The most negative sentence in your post had a normalized negativity score of -0.026999999999999996:
"After starting the discord server and inviting @DNews he joined right away."
As good as this idea sound, I find the math totally arguable.
I was wondering about that to
Posted using Partiko iOS
If you're saying that you feel that positivity and negativity are subjective, I agree. This has been a topic of discussion in many other posts. At the end of the day, the bot uses an open source list of words tagged as 'positive' or 'negative' . Words like 'bad', 'ugly', 'evil' are negative, while words like 'happy', 'fun', 'good' are positive. These lists are subjective for sure, but the output of the bot is deterministic given an input list.
It is an interesting idea and I can see possible use cases
Posted using Partiko iOS