README

Jordan Sanders
CSCE 5200
Final Project

Programs:
download.py
make_model.py
query.py
utils.py

Other files:
stop_words.txt
tfidf.pickle


To perform a query:
python query.py

To download 3000 webpages:
python download.py

To create the vector space model:
python make_model.py


Notes:
I have NLTK and one of its dependencies imported into the main project directory (Six)
tf and idf values have been pre-computed in the tfidf.pickle file for quick run time.


Descriptions:
query.py:
  This program takes a query argument, preprocesses it, and then calculates the cosine similarity between
  the query and all documents in the model. It then ranks them in descending order of similarity and outputs
  the top 5 URLs with the highest similarity to the query. It then prompts the  user to continue or end.

utils.py:
  This file contains all functions written in this course. Specifically written for this
  project was the checkURL and fixURL functions.

download.py:
  This program uses the urllib2 and bs4 libraries to download webpages and parse them
  for all links. Links are then added to a queue from the queue library using breadth first search.
  All links already visited are not added to the queue. Downloaded pages are stripped of
  all markup text and written to a cPickle binary file to be used for creating the model later.

make_model.py:
  This program opens the pickled webpage text files and runs the tfidf_text function
  that creates the vector space model using a dictionary that is then stored in a pickle
  file called tfidf.

tfidf.pickle is the tf and idf dictionaries pre-computed for quick run time. Delete for re-compute