Document Classification

This is an implementation of document classification using four algorithms: Multinomial Naive Bayes, Support Vector Machines, Random Forests and Logistic Regression.

Validation Sore is calculated as ( #correct - #incorrect )/( total )
SVM has shown to have the highest validation score of 94%

Requirements:

1. Python 2 or 3
2. sklearn python module
3. nllk python module and download stop words beforehand if possible
4. pickle python module
5. sys python module
6. time python module
7. pathlib python module
8. future python module

Variables

ALL of these variables can be found in the main() function

1. cachedStopWords = False - list of stop words to remove.
   you can define your own stop words. if "False" it will download from NLTK and use stopwords provided from them.
   cachedStopwords = stopwords.words("english")
2. training_file = "trainingdata.txt" - this is the path of training file.
3. testing_file = "testingdata.txt" - this is the path of the testing file.
4. save_model_file = "document_classifier.sav" - this is the file in which the model wll be saved in.
5. save_vectorizer_file = "tfidf_vectorizer.sav" - path of the file to save the vectorizer model in (should be a .sav file)
6. load_model_status = True - if true it will load the model from the file with the name stored in the variable load_model_from, if False it will create a new model.
7. Load_model_from = "document_classifier.sav" - this is the path of the file from which the model will be loaded from.
8. load_vectorizer_from = "tfidf_vectorizer.sav" - path of the file to load the vectorizer from. if the file does not exist, it will use the tfidf vectoizer from scikit-learn by default.
9. validation_size = 0.25 - this is the split ratio for training vs validation data.
   -> size(validation_data) = validation_size * size(total_data)
10. classifier_to_use = "svm" - this is the classifier to use.
    Choose from: "Multinomialnb", "svm", "random_forest" and "logistic_regression"

functions

score() function that will calculate the score as 100*(#correct-#incorrect)/(T)

load_test_file() function to load the testing file data given the testing_file would be a certain format

help() function to display help please note for this function to work you will need the help.txt (path stored in the variable help_file)

"document_classifier" is a class with the following member functions: init(), train(), predict() and save_model()

To run

python document_classification.py
-> normal

python document_classification.py -h
-> enables help

python document_classification.py -help
-> enables help

Training file should of the following format:

• The first line contains the number of lines that will follow.
• Each following line will contain a number (1-8), which is the category number. The number will be followed by a space then some space seperated words which is the processed document.

Testing file should of the following format

• The first line in the input file will contain T the number of documents.
• T lines will follow each containing a series of space seperated words which represents the processed document.