Sarcasm-Detection

Trains several classifiers on a dataset obtained from Reddit. Features inclue unigrams, bigrams, trigrams, and word embeddings. Word embeddings obtained via pretrained word2vec model.