Data

The data used in this visualization comes from Infegy, a consumer intelligence platform that gathers, analyzes, and presents information from billions of social media posts. The list of bands was obtained from a publicly available source on GitHub.

Visualization
The process of the visualization involves searching for posts about the Red Hot Chili Peppers (RHCP) and analyzing the data to find the top 50 bands mentioned in these posts. For each of these 50 bands, the process is repeated to find the other bands they are connected with. The connection between two bands is represented by an edge, and the strength of the edge reflects how often they are mentioned together.
The results show that The Beatles are at the center of all the connections, confirming their strong influence. Additionally, the data reveals that the Rock cluster is separated from the Alternative and Pop clusters.

Tools
The tools used for this visualization include Python and its libraries Pandas and PyVis.

This is good feedback, thanks!

Yeah, that was a tricky bit. 538 often has longer type polls (e.g. this poll is valid from 10/1- through 10/5). To handle those, I just took the beginning of the date range.

Data:
Social Media data sourced from Infegy, a consumer intelligence platform that collects, analyzes, and visualizes billions of social media posts.
Polling average from FiveThirtyEight’s summary of Joe Biden’s approval rating, available here: FiveThirtyEight- How Popular/Unpopular is Joe Biden?

Explanation:

This visualization tracks the aggregate average of online sentiment for posts across the internet discussing Joe Biden from January 1, 2021 through the present (n=74,409,409 individual posts). I compared that with the moving average from FiveThirtyEight’s How Popular/Unpopular is Joe Biden?
The original idea for this graph came from a Politico article that said that after the Taliban’s takeover of Afghanistan, Joe Biden’s popularity began to slip (Politico - Biden's Afghanistan Withdrawal Anniversary).
Interestingly, it looks like aggregate sentiment data form social media bears that out to some degree. Biden’s sentiment online drops as the Taliban began their advance. That drop in sentiment is matched with a drop FiveThirtyEight’s polling.
Overall though, it appears that social media data is more reactive, but less durable than traditional polling. Negative/Positive stories break in the news, which causes sentiment to go up or down.

Tools:
Python (Pandas and Matplotlib)

I'm getting a definite blood spatter vibe from this.

Do you like it? We're using currently using Hubspot.