TwitMiner

TwitMiner

TwitMiner Are you a social media addict? Are you a follower of celebrity pages and accounts? Are you interested in what’s happening around you? Does challenging problems set adrenalin racing through your veins? Then this is the right place to be in.

The Computer Science and Automation department of Indian Institute of Science, Bangalore presents the first of its kind Machine Learning contest in India: Develop a prediction algorithm by learning from a collection of texts(tweets) hosted by Twitter.com.

The challenge is to predict whether a particular tweet text can be classified to a category of ‘Politics’ or ‘Sports’. This is a simple binary classification supervised learning problem, where you will have a training set consisting of tweettexts and category labels and you have to come up with an algorithm which predicts correct labels for another set of tweettexts. You will be judged on how accurate your algorithm is.

Since Twitter.com API rules forbids distribution of tweet texts to any third party, we will not be sharing the tweet texts for the training or test data. We’ll be sharing only the tweetids, userids and hand-curated labels as the dataset. In addition to that, for the convenience of the participants, we’ll also be providing a script which will help them to download the required tweet texts.

Prizes:

First Prize: Internship/PPI offers at Amazon.com for Machine Learning SDE or Machine Learning Scientist depending upon the candidate.
Second Prize: Prizes worth 10000 INR.
Third Prize: Prizes worth 6000 INR.

Please note: Internship offer (without any interviews) will be offered to non-final year students and Pre-Placement Interview call(the written and coding rounds waived off) will be offered to final year students. The contest is open to undergraduate as well as graduate and PhD students. Working professionals(maximum work experience of 7 years) are also encouraged to participate.