Sentiment Analysis using Doc2Vec uses doc2vec as part of the sentiment analysis which might be worth a shot. doc2vec is a way of creating a single vector for a document (or any collection of words). Since its unsupervised it could potentially be trained on many manuscripts. But gensim might not scale so well. It might be more useful to use a TensorFlow implementation with pre-trained word vectors.
(Personally though I am slightly sceptical that an unsupervised method will generate a good document vector for sentiment analysis - but I might well be wrong)
Sentiment Analysis using Doc2Vec uses doc2vec as part of the sentiment analysis which might be worth a shot. doc2vec is a way of creating a single vector for a document (or any collection of words). Since its unsupervised it could potentially be trained on many manuscripts. But gensim might not scale so well. It might be more useful to use a TensorFlow implementation with pre-trained word vectors.
(Personally though I am slightly sceptical that an unsupervised method will generate a good document vector for sentiment analysis - but I might well be wrong)