Tracking sentiments on Twitter over time

The sentiment timeline on Twitter Sentiment is a very useful feature that allows you to track sentiments towards a particular query term over time. By default, Twitter Sentiment tracks sentiments for popular queries like "Google", "iPad", "Obama", etc. But you can add your own custom queries to track - For instance, I've been tracking the query "Indian Cricket Team" since May 17th (i.e. after the team's T20 World Cup debacle):



The timeline is helpful in two ways. First, it gives you an indication of how much buzz there is around the topic, i.e. how much it is being talked (tweeted) about. Secondly, it gives you an idea of the sentiment towards the topic.

As the graph above shows*, there was still some residual anger towards the team for a couple of days after the end of the world cup, but then it settled down for a while. On May 27 (PDT) however, there was a surge of negative tweets about the team, following their loss to Zimbabwe. Sentiments started improving (the positive line climbing up and the negative one coming down) on May 29 after their victory against Sri Lanka... but soon came the announcement about the Indian team not being sent to the Asiad and sentiments started being negative again on May 31. And as one can expect, it only grew worse after their second defeat (and a very embarrassing one) to Zimbabwe on June 2.

Update (12/28/2011): We temporarily disabled this feature. If you would like to track queries over time, please let us know by describing your use case, so that we know how to prioritize this feature.

Google Gadget for Twitter Sentiment

We now have a gadget for tracking Twitter Sentiment.

Try it! You have two options:
1. Add the gadget to your iGoogle page
2. Embed the gadget on your webpage [Video tutorial]
Please let us know what other features you want through our feedback form.

Index for Stanford Machine Learning Lectures

Stanford has posted videos to its machine learning class, taught by Andrew Ng. It's a very good class. Unfortunately, the videos don't have an index so it's difficult to determine what each lecture is about.

As I watch these videos, I started indexing the contents here:
http://spreadsheets.google.com/pub?key=t-Noh8oUPc_yHGOBvAJldbQ&output=html

If you want to contribute, you can edit this spreadsheet:
http://spreadsheets.google.com/ccc?key=t-Noh8oUPc_yHGOBvAJldbQ&hl=en

Historical Data Now Available

You can now get historical data on tweets. Here's what it looks like:

Alternatively, you can try out these queries:
- seth godin
- iphone

- time warner

How to track your own queries:

1. Click on "Save this search" next to the search box (highlighted in screenshot above).

2. Login with a Google Account (We don't see your password. Ever.)

3. We'll start tracking the query over time.

A few notes:
- It may take a few days for your query to gather stats.

- This is still an pre-alpha feature. Please let us know if you see any problems by leaving feedback here.

- We only support English queries at the moment.

How many people say "huuuuuuuuuuungry" on Twitter?

Here's a fun example that shows the uniqueness of the Twitter language model. Try typing in the word "hungry", with an arbitrary number of u's. Seriously, try it:

huuuungry - 15 results in the last day
huuuuuuuuuuungry - 8 results in the last 10 days (screenshot below)
huuuuuuuuuuuuuuuuuuuuungry - 1 result in the last 4 days



















The publicly available lexicons (like the Subjectivity Lexicon from MPQA) don't work well with Twitter because the language model is so unique.