The file CodeSource.rar contain an archive with main file that 
main process the Youtube corpus about net-activism and whistleblowing 
see https://doi.org/10.5281/zenodo.5824627 


the file _pipeline.txt describe different steps : 
- download, 
- storage in a Mongo database, 
- splitting id(s), 
- filtering collection, 
- creationg of collection with only text and sentences, 
- linguistic feature extraction , 
- features extraction, 
- clustering

the sub-directory called ExtractYoutube is java code for getting transcription from id


To run source code require, hadoop server and lots of libraries such as :

1.  OS Linux
2.  MongoDB        https://www.mongodb.com/download-center?filter=enterprise#enterprise
3.  JAVA 1.8 or later
4.  Python 2.7 or later
5.  Maltparser     http://www.maltparser.org/download.html
6.  Stanford NER   http://nlp.stanford.edu/software/CRF-NER.shtml
7.  TreeTagger     http://www.cis.uni-muenchen.de/~schmid/tools/TreeTagger/
8.  SRILM          http://www.speech.sri.com/projects/srilm/download.html
9.  MorphSegmenter https://www-i6.informatik.rwth-aachen.de/~mansour/MorphSegmenter/
# About python
1.  langdetect-1.0.6       https://pypi.python.org/pypi/langdetect?
1.  Morfessor-2.0.2alpha3  https://pypi.python.org/pypi/Morfessor/2.0.2alpha3
2.  nltk-3.2.1             https://pypi.python.org/pypi/nltk/3.2.1
3.  numpy-1.9.3            https://pypi.python.org/pypi/numpy/1.11.2rc1
4.  polyglot-master        https://pypi.python.org/pypi/polyglot/16.7.4
5.  pycld2-0.31            https://pypi.python.org/pypi/pycld2/0.31
6.  PyICU-1.9.3            https://pypi.python.org/pypi/PyICU/1.9.3
7.  pymongo-3.3.0          https://pypi.python.org/pypi/pymongo/3.3.0
8.  six-1.10.0             https://pypi.python.org/pypi/six/1.10.0
9.  wheel-0.29.0           https://pypi.python.org/pypi/wheel/0.30.0a0
10. JIEBA                  https://pypi.python.org/pypi/jieba/