I'm working on a python application that analyze the market feeling about a particular stock. I'm looking forward to open source after the basic part is done.

1)It finds and prepares database of tweets about public companies by searching tweets containing Company's name or ticker symbol.

2)Tags them as positive, negative or neutral comment.

3)Prepare a report depicting market sentiments about a particular stock over a period of time.

4)Compare it with how stock is actually doing in that time period and after that.

It is an experimental project to check if we can predict how the stock will fare by measuring the market sentiments toward it and if it is reliable or not.

Wow, that sounds great! Make sure to share the results with everyone here on HN!

Interesting - how are you doing the sentiment analysis?

1)First few (say 1k or as many one wants) will be done by manually tagging the tweets as +ve, -ve or neutral, through a web or command line interface.

2)Then, I will use Google Prediction API.

3)Based on this data we can get a graph depicting the rise and fall of overall sentiment for a particular stock over a period of time.

I finally got my webcam to work on Windows with Processing (java variant used for image...processing. worst name ever) so I'm gonna build a "security cam" application for my room. Simple enough - constantly refresh the cam feed, check for significant changes in the image. I think I might try and add facial recognition for people who live in my house so if the camera catches them a script will send them an email or text saying "Stay out of my stuff!"

Lol. That's nice. How are you gonna do face recognition?

Using this - http://developers.face.com/

??