*JESS3
AnOrangeAmerica.com
Project powered by: Tropicana
Site Name:
Freshly Squeezed Election Tweets
"We're not red. We're not blue. We're 100% orange."
Site URL:
Launched:
This morning, when the polls opened!
Tagline:
"A place where you can compare oranges to oranges."
Overview of project:
In an online space where lots of information is streaming in, users need help making sense of what is being said and what it means
The site itself is an abstract visualization of the aggregate conversation on the popular social media platform, Twitter. It shows frequency and context of election-related terms, live as they happen. Consider it a living information graphic: the picture changes over time as the conversation develops over time.
The site will pull a continuous stream of tweets mentioning Obama and McCain. Each tweet is examined for certain key words and phrases.
We have a pre-determined list, but we will be looking out for other words appearing with significant regularity and adding them to the tool.
The data is represented as a series of interconnected half-circles. The bigger the "bubble" the more frequently the term is being used. Each bubble is colored red and blue: if the bubble is mostly red that means the term correlates more strongly with mentions of John McCain, and vice versa: the more blue, the more it's associated with Barack Obama.
Keyword bubbles are themselves connected by half-circles (solid colors only), showing relationships between them. The stronger the connection, the thicker the line.
You can make it show what you want: show only tweets about Obama or McCain, about both, or about one but not the other. If you want to compare the current conversational landscape against a full day's worth of tweets, or even three days, you can do that, too.
You can embed it and go full screen!
Technical Details
How does it work?
We take a sample from Twitter every 30 seconds and analyze them in 50-result batches for associations and term matches. They accumulate for 5 minutes and then we flush sample aggregates to the database. So the database has samples from when it started through present in 5-minute granularity. As new terms trend, they begin to populate on the X-axis.
What's under the hood?
The system back end will be implemented using a Java-based stack and PostgreSQL RDBMS. The presentation will be implemented using Flash targeted to Player 9, standards-compliant XHTML/CSS targeted to modern browser versions with significant market share (Safari 2+, Firefox 2+, IE 6+).
Source: www.AnOrangeAmerica.com
AnOrangeAmerica.com
Project powered by: Tropicana
Site Name:
Freshly Squeezed Election Tweets
"We're not red. We're not blue. We're 100% orange."
Site URL:
Launched:
This morning, when the polls opened!
Tagline:
"A place where you can compare oranges to oranges."
Overview of project:
In an online space where lots of information is streaming in, users need help making sense of what is being said and what it means
The site itself is an abstract visualization of the aggregate conversation on the popular social media platform, Twitter. It shows frequency and context of election-related terms, live as they happen. Consider it a living information graphic: the picture changes over time as the conversation develops over time.
The site will pull a continuous stream of tweets mentioning Obama and McCain. Each tweet is examined for certain key words and phrases.
We have a pre-determined list, but we will be looking out for other words appearing with significant regularity and adding them to the tool.
The data is represented as a series of interconnected half-circles. The bigger the "bubble" the more frequently the term is being used. Each bubble is colored red and blue: if the bubble is mostly red that means the term correlates more strongly with mentions of John McCain, and vice versa: the more blue, the more it's associated with Barack Obama.
Keyword bubbles are themselves connected by half-circles (solid colors only), showing relationships between them. The stronger the connection, the thicker the line.
You can make it show what you want: show only tweets about Obama or McCain, about both, or about one but not the other. If you want to compare the current conversational landscape against a full day's worth of tweets, or even three days, you can do that, too.
You can embed it and go full screen!
Technical Details
How does it work?
We take a sample from Twitter every 30 seconds and analyze them in 50-result batches for associations and term matches. They accumulate for 5 minutes and then we flush sample aggregates to the database. So the database has samples from when it started through present in 5-minute granularity. As new terms trend, they begin to populate on the X-axis.
What's under the hood?
The system back end will be implemented using a Java-based stack and PostgreSQL RDBMS. The presentation will be implemented using Flash targeted to Player 9, standards-compliant XHTML/CSS targeted to modern browser versions with significant market share (Safari 2+, Firefox 2+, IE 6+).
Source: www.AnOrangeAmerica.com