added script to download and prepare sentiment data

#!/bin/bash

# file: prepare_data.sh

#

# Program: Downloads and prepares data containing the sentiment dataset.

#

# Author: Thomas Huetter

# create target folder and change into it

mkdir sentiment

cd sentiment

# download the data files

wget https://nlp.stanford.edu/sentiment/trainDevTestTrees_PTB.zip

# unzip the data into folder original_data

unzip trainDevTestTrees_PTB.zip

# remove zip file

rm -rf trainDevTestTrees_PTB.zip

# change to unzipped folder

cd trees

# prepare data for file L.trees

# convert dev.txt and train.txt into UTF-8 format |   replace ( by {    |   replace ) by }    | sort by number of nodes (equivalent to number of "{")

iconv -f ISO-8859-1 -t "UTF-8" dev.txt train.txt  | sed -e 's/(/{/g' | sed -e 's/)/}/g' | awk '{print gsub("{","{"), $0}' | sort -n  | cut -d' ' -f2- > ../sentiment.bracket

# go back to the folder

cd ..

 No newline at end of file