Use nlp techniques to observe the meannings of words change with time. Also used for cmsc723 project. Instead of training word vectors on co occurance can you use a loss function based upon year distribution that word was used in?
Models found in '/preprocess/'.