- NumPy = 1.15.3
- stop_words
- nltk
read datasets in /data/ by 'read_raw_text.py' and read datasets in /souce/ by 'read_semeval.py'
extract dependency relations by 'dependency_parser.py' apply Double Propagation by 'double_propagation.py'
run 'nmf.py' to cluster aspects and predict the implicit aspect