A 2015 CS 410 - Text Information System project
Twitter promotes high interactions among users. 500 million tweets are produced per day and roughly 10% of these are questions tweeted by users.
In the first phase of this project we aim to identify such tweets. We focus on
- Classifying tweets into a question and non-question.
- Classifying question tweets into machine-answerable or not.
#Technique
Towards this end we implemented a technique that involvs 4 steps
- Data gathering
- Feature extraction
- Feature selection
- Training classifiers
Our results have been encouraging so far.