Machine learning models for sentiment analysis on tweets.
$ python main.py <name_experiment>
Just add a classifier and a feature extractor classes in the directory experiments
.
Refer to this webpage.
Depending on the experiments, one may need: pandas, sklearn, nltk, keras, gensim, numpy, re, pytorch.
We use stopwords.words('english') from nltk.corpus. To get this stopwords corpus, use the NLTK Downloader. Open a Python console and do the following:
>>> import nltk
>>> nltk.download()
For the RecursiveNN and CNN parts, we use pretrained embeddings. One can get these files (.zip) at https://nlp.stanford.edu/projects/glove/, unzip them and put them in the experiments folder.