Final Project for CS 5340 / 6340, Coreference Resolution

###Authors

• Adam Walz

• Charmaine Keck

This project is compressed in a .tar.gz file due to the handin utility not allowing students to hand in a file structure.

To uncompress run

tar -xvzf coref.tar.gz
cd final

A Makefile was added to allow the easy configuration and running of this project. Its purpose is to point to a specific python installation (python 2.7.3) as well as datasets used for the python natural language toolkit (nltk). Running the make command will run the entire project and place output files in the desired output directory.

As arguments, pass make the paths to the listfile and output directory

make ARGS="listfile output_dir"

###Example Given that the listfile is at: /home/cs5340/Project/DryRun/testset1/filelist2

And the output directory is: /home/cs5340/Project/DryRun/Results/output/

The make command would be make ARGS="/home/cs5340/Project/DryRun/testset1/filelist2 /home/cs5340/Project/DryRun/Results/output/"

This both configures and runs the project, and places the results in the output directory

This is not recommended due to different versions of python on the cade machines

Running the help command python coref.py -h prints

Coreference Resolution Engine

positional arguments:
listfile              File containing file path strings
responsedir           Path to output directory

optional arguments:
-h, --help            show this help message and exit
-v, --verbose         increase output verbosity
-t, --test            run doctests only
-H HOST, --host HOST  Host address of stanford corenlp server

Satisfies final project for CS 5340/6340, University of Utah 
AUTHORS: Adam Walz (walz), Charmaine Keck (ckeck)

For more information, please see project
description http://www.eng.utah.edu/~cs5340/project/project.pdf