Twitrratr is a StartupWeekend project that measures public opinion on Twitter. It allows you to measure general opinions on a topic by entering a keyword and cross-checking it against a list of pre-compiled adjectives. There are lists for both positive and negative adjectives; anything that doesn’t match falls in the neutral category.

The real motivation behind the project was to determine Twitter opinion of both candidates in the 2008 US Presidential Election. Since then, it’s spread to literally any keyword you can think of that has been mentioned on Twitter. More tweets on a keyword leads to greater accuracy, as always.

The Good

Twittrratr does a pretty good job of measuring positive opinion. The adjective list for positive opinions works better than the negative opinions list does. It’s also pretty interesting to quantify opinion into categories to see how people feel about something. It’s generally fairly accurate, but does suffer from some glaring issues with successfully determining opinion.

The Bad

The keyword lists that drive the entire categorization of tweets is fundamentally flawed. It’s nearly impossible to determine a person’s opinion based on keywords. For instance, when searching for mint.com, you’ll find that @Teanna said “Mint.com is pretty awesome.”. Somehow, this got lumped into the neutral category. This is just one case of many instances. Another shining example comes from @integrateit, who said:

From Bitbucket Mint vs. Quicken, and the importance of a good UI: Mint.com’s Aaron P.. http://tinyurl.com/69ydpn

This user didn’t say that Mint.com is good, he just mentions the word good in reference to a UI design.

The Verdict

I think Twitrratr is, at the very least, a good way to find out what users are saying about your brand by attempting to quantify their tweets into categories. It’s not a very mature system and could be improved using better algorithms or analyzing tweets in more depth. Even though it’s not completely accurate, it’s still a fun tool to check out every once in a while.

Update: Twitrratr has won the Mashup of the Day Award for November 13, 2008.

If you ask anyone using Twitter about the depth and breadth of third-party Twitter applications, you’re likely to receive a positive response. The appeal of Twitter is so great to third-party developers that they all want to spend their time writing applications for the platform. The number of paths a developer could travel is mind numbing - there are status updaters, stat trackers, and follower counters, just to name a few. What I’m most intrigued by is the status updaters.

As a developer, the goal is not just to create an appealing, functional application, but to also have a solid user base. In a highly saturated market, it’s critical that the application does something better than the competition. So how does a developer know how many people are using their app? One new option is to use the Twitter Update Source Tracker (Tracker for short).

When a developer creates and deploys an application, it’s beneficial to them to know the amount of use it gets. Tracker can determine the number of updates sent by any given source and figure out what percentage share that service has. Devs can also chart their progress by using Tracker’s Trends feature. This allows them to see what kind of progress they’re making on a daily, weekly, or (eventually) monthly basis.

Tracker’s Trends can be used as a success gauge for a third-party application. Success comes in two flavors - moving up the chart and staying toward the top of the chart. Movement up or down the chart is shown with up and down arrows and numbers. Other users can also come to the site to see which applications are hot. Surely some will try the applications that are moving up on the charts to see what the buzz is. This is another way that Tracker can help devs - by giving them free publicity. Is an application rocketing up the charts, or dropping like a rock? Now you’ll know.