tweetfeel/biz gives the Inside Scoop on Balloon Boy Sentiment

Monday, November 16th, 2009

Balloon Boy quickly faded into history once the story was discovered to be a hoax. But, the tall tale of a boy set adrift in a balloon provided much fodder for the twitter world. Let’s go on a Balloon Boy journey ourselves, but this time through the eyes of tweetfeel/biz.

We begin our journey on October 15, 2009. The news broke and Twitter was immediately alight with Balloon Boy tweets. Based on the two week period from October 15 to October 29, the blue line shows that 31% of tweets occurred on Day 1, dropped to half that amount by Day 2, and almost disappeared ten days after that. Day 1 tweets hovered at the 50/50 positive negative line (green line) as some people immediately felt it was a hoax while other people expressed concern for the safety of the phantom boy. But, the tide shifted just a few days later as it became widely known that Balloon Boy was a hoax. At this point, the sentiment, among those few people still tweeting, became overwhelmingly positive.

But what were people really thinking? Why did emotions become so positive for a situation that caused dire panic among emergency rescue personnel? A bit of digging into some tweetfeel/biz themes helps to build this story. It turns out that people weren’t expressing their positive sentiments in the sense of being happy about the situation, but rather they were expressing happiness by way of humor. In fact, about 65% of tweets focused on the humour of the situation! People were particularly interested in the fact that the media picked up on the tale even though it was a hoax.

• LOL I love how Balloon Boy is a trending topic.
• Haha, #balloonboy was in the house the whole time. American news just got pranked.
• LMAO balloon boy. Nicely done.

Not everyone thought it was so funny though. Another 21% of people were angry about the situation for several reasons. They were upset that they had been deceived themselves, that people’s time had been wasted worrying and attempting to save a child, and that the parents had simply behaved in a disrespectful manner towards everyone including their own children. Another 13% of tweets expressed sadness, particularly in the early days, about the situation assuming it was real, and then the sadness subsequently turned to disappointment.

• BOOOOOOOO!!!!!!!!! I hate you Balloon Boy. Way to let me down. You trickster.
• i hate #balloonboy as much as I hated #eliangonzales
• Eff you balloon boy!
• Awww the 911 call made me cry balloonboy
• The Colorado Balloon Boy family is pathetic !
• It landed, and nobody was inside. This is very sad. #balloonboy

If you’re still curious about the incident, Wikipedia has an article that will answer all your questions. If you’re curious about tweetfeel/biz, give it a free testdrive yourself over at tweetfeel.com/biz.

Come See Us at TWRTCON and 140

Sunday, October 18th, 2009

tweetfeel/biz is ready to leave the nest and we’re ready to announce it at two upcoming conferences.

First off is TWTRCON in DC on October 22 where we are a sponsor and will be running demos of tweetfeel/biz. Then, on October 27, we will be at the 140 conference in LA also sponsoring and running demos for attendees. Both Jean and Tessie will be there signing autographs as well so be sure to stop at our booth! They’d also love to give you a personal demonstration of how you can use tweetfeel/biz to help your business. See you there!

Sentiment Analysis 101

Friday, July 17th, 2009

tweetfeel gives you a taste of it, but really, what is sentiment analysis all about?

At it’s most basic level, sentiment analysis involves reviewing messages or conversations and evaluating the writer’s opinion towards the topic. For instance, someone who tweets a message such as “I like Chuck Norris” is telling people they have a positive opinion towards Chuck Norris. On the other hand, someone who writes “Chuck Norris sucks” clearly has a negative opinion. After assembling all of the messages that mention Chuck Norris, one can easily bucket them into messages with positive opinions and messages with negative opinions.

But, the easy part isn’t so easy. First, one needs to determine which sentiments are positive or negative. Obviously, we’re talking automated sentiment analysis so we need some solid indicators for positive opinions such as words like happy, love, or delightful. Solid indicators for negative opinions would be words such as hate, stupid, or ugly. Simply coming up with that list is difficult enough, but some words just aren’t so easy to assign to buckets. For instance, is “Way to go” positive or negative? People often use this phrase in a positive way but in recent years, it has become a very sarcastic remark that one uses in a negative fashion. The written word is full of words and phrases that have contradictory, ambiguous, or sarcastic meanings. Humans can only catch about 85% of those which means it’s pretty much impossible for an automated process to catch all of them either.

Another problem with bucketing messages is that people don’t think linearly. If I say “I love Chuck Norris and football sucks,” it’s clear to people that I’ve messaged two distinct opinions about two distinct topics. Once you start getting into more complicated grammar though, it can become impossible to tell which topic was rated which way. Automated evaluations of the message have a much harder time differentiating the two. It’s a topic of great interest to academics and eventually, we’ll figure it out.

In the end though, it’s not about individual messages. It’s not about me and what I have to say. It doesn’t matter that your uncle Bob is always wrong and that your Aunt Mary doesn’t know who Chuck Norris is. It doesn’t matter that 5% or 10% of the messages are in the wrong bucket. What matters is the collective wisdom, the wisdom that comes from large sample sizes. When you average opinions across hundreds or thousands of people, the final answer is usually the right one.

TweetFeel Testimonials

Wednesday, July 15th, 2009

Well, tweetfeel officially launched yesterday and we’ve had an amazing response! We are delighted to hear that so many people are having fun checking it out. In fact, so many people have been checking it out that we hit our Twitter API call rates and haven’t been able to run everybody’s search! Don’t worry, though. We’re speeded things up from our end so you should see some improvements soon!

We’ve been reading all the comments people are making about tweetfeel.  You’re right, it’s not perfect nor will it ever be perfect. But, it’s fun, it’s pretty quick, it’s pretty accurate, and it’s free. How can anyone resist that?

Below are just a few testimonials from our friends. Thanks guys!

1. bootdisk If you’ve not checked out http://www.tweetfeel.com/ yet – have a look, interesting way of measuring ‘feelings’ towards people’s tweets!
2. chrish10 @VolumeGroup im impressed with #tweetfeel, looks very good and extremely useful
3. LenKendall Have you checked out Tweetfeel.com? It’s a very cool tweet sentiment application. Simple/Pretty interface. http://is.gd/1ytSQ
4. zoodigital Nice twitter tool – http://www.tweetfeel.com
5. Legend_Kiiller Crazy! RT @mashable: TweetFeel: Real-Time Sentiment Search for Twitter – http://bit.ly/JpBsjabout
6. wearelondon tweetfeel.com cool error msg “We have failed you and we are sorry.Try another search term & we promise, we’ll do better…” (@AndrewGrill)
7. hridayramshenoy checking out TweetFeel. Damn good idea.
8. vishnugopal TweetFeel – Really cool. Try Windows Vista or Macbook pro http://tr.im/sdwU (via @brupm)

TweetFeel: What’s the Twitter feeling?

Friday, July 10th, 2009

We’ve been dying to share with you our great new app called tweetfeel. This new application scours Twitter for Tweets about brands of your choice and shows you how positively or negatively Twitter users feel about it. Whether it’s Dell or Poptarts or Madonna, we’ll show you a sample of recent tweets as well as an overall rating for the brand. It’s a quick way to get a feel for what people are thinking!

tweetfeel uses an algorithm that is more complicated than simply counting happy faces and sad faces. Because only a small sample of people use emoticons , apps doing it that way miss out on the majority of people who talk about brands using words. This means that tweetfeel produces results that are more accurate than other applications doing a similar thing

At the time of this post, Twitter tweets were 63% positive, Dell tweets were 53% negative, and Poptart tweets were 92% positive. Try out your favorite brands to see how they rank!

Follow tweetfeeldotcom on Twitter and we’ll share some results with you every day.