Web information systems - Authorship profiling

• DELIVERABLES/ contains

  • automatically labeled data (separated in to sets, one for testing(1/3) the rest for training(2/3))
  • manually labled data
  • Report

• classifiers/ contains the classifiers used to fill in the information about our profiles

  • most are rule base and implement by hand
  • some use external APIs such as ageanalyzer.com or lymbix for age and sentiment detection

• data/ contains all our data collections (either in .)son or .shelve)

• models is what interacts with the .shelve files in data/ and structure the information

To rebuild the dataset, you need to run python buildDataSetsForServer_new.py. It will generate all the data and put it inside data/test_data/, than you can copy this new data in DELIVERABLES/auto_labeled_data/

• python-twitter: https://github.com/bear/python-twitter
• lymbix https://github.com/lymbix/Python-wrapper
• PyQuery http://pythonhosted.org/pyquery/api.html
• nltk http://nltk.org/
• ageanalyzer.com