The Twitter Sentiment Index
The Twitter Sentiment Index is an index created to analyze the realtime public sentiment of 2020 Presidential Candidates using Twitter.
#4 Product of The Day - ProductHunt
The Future of Public Opinion Polls Will Move To A Similar Technique.
Why?
Most News Networks have a political bias. Using them as your only frame of reference will give you a skewed view of the world.
I created this Index because I wanted to see how the world (Twitter Users) actually felt about Presidential Candidates.
Technique
Every 15 minutes, The TSIndex compiles a list of 100 of the latest tweets related to a specific presidential candidate. Using a variety of AI Apis & Libraries, each tweet is analyze to generate a Sentiment Score from 0 to 1 (0 = Negative 1 = Positive). We calculate a average score from the 100 tweets and save the value. We later use a moving average formula to smooth out the raw data and generate a readable graph.
This technique is not perfect but there are some insights we can take from the data.
There are some downsides to this technique which should not be forgotten. We should question if Twitter users are an accurate sample of US Voters.
Though most tweets compiled were of US Based users, there may be a significant amount of non-American users tweeting about presidential candidates.