mktweets is an attempt to construct a gender classifier for Macedonian twitter users.

First, the following dependencies need to be satisfied:

• nltk
• python-twitter
• mongodb + pymongo
• guess_language from https://bitbucket.org/spirit/guess_language
• numpy
• scipy
• matplotlib
• pylab
• scikit_learn

In order to crawl tweets and make a dataset to learn from, put the twitter api key in the module folder named twitterkey, change the mongodb ip:port as well as the starting username in crawl/crawl.py, run the mongodb server and then start crawl/crawl.py

In order to train the module, change the number of records used for training/test in train/train.py as well as hand-tag them in the same file, and then run it. Output is in console as a comparison chart between different methods.