Authorship identification has been a very important and practical problem in Natural Language Processing. The problem is to identify the author of a document from a given list of possible authors. A large amount of work exists on this problem in literature. We develop ideas based on this work in order to build our own model for authorship identification. We also take a model from this work as a baseline for comparing the results. Our model for the task is a text classifier based on logistic regression which includes n-grams, style markers and document finger-printing as features.
Reuter_50_50 is the dataset used. It is present in directories training/ , testing/ and all. It contains 50 text file for 50 authors. Each text file contains several lines for that author.
python with common ML and NLP libraries like Scikit-learn,Theano,Nltk etc.
learner.py is the main file . run it to see the output.
coming soon
- Fork it!
- Create your feature branch:
git checkout -b my-new-feature - Commit your changes:
git commit -am 'Add some feature' - Push to the branch:
git push origin my-new-feature - Submit a pull request :D