You have the following list of dataset to help you managing it :

• directory_creator.sh : create directories automaticly

	echo "Usage : directory_creator.sh -[h|c|y|d] (name) (year)"
	echo "-h : affiche l'aide et quitte"
	echo "-c : creer le pays 'name'"
	echo "-y : ajoute l'annee 'year' pour le pays 'name'"
	echo "-d : ajoute un debat dans le pays 'name' pour l'annee 'year'"

• create_aam_as_files.py : create .aam and .as files (need to use directory_creator before) See below for informations about the strcuture of infos.xml

The dataset is group by country and year of debates following this tree:

dataset/
	country_name/
		year/
			num_of_debate/
				annotated/
					ac-aa/
					debate.aam
					debate.as
				brut/
					x.txt
					...
				full/
					full.txt
				reactions/
					x.txt
					...
				segmented/
					x.txt
					...
				infos.xml

Descriptions

• full.txt contains the full script of the debate
• brut/ contains the brut script cut in topics(questions) numbered from 1 to x
• segmented/ contains the segmented script cut in topics(questions) numbered from 1 to x with '&' for split character (need to add auto generate method for this)
• reactions/ same format as brut but contains also the reactions from the persons (format will be changed in the future for easier use and annotation [glozz])
• annotated/ contains the files for glozz annoted and the parameters file (need to add auto generate method for this)
• infos.xml contains data about the debates

(need to add script for auto generate glozz files and segmented files and annoted reations easier)

The file need absolutly to respect this format :

<?xml version="1.0" encoding="UTF-8" standalone="no"?>
<informations>
<debateNum>1</debateNum>
<country>usa</country>
<language>en</language>
<numberQuestion>9</numberQuestion>
<date>09/26/16</date>
<participant>
<presentator>HOLT</presentator>
<candidate>CLINTON</candidate>
<candidate>TRUMP</candidate>
</participant>
</informations>

The script need to be like below:

NAME: text

NAME: text