Java Deep Learning Projects

Preparing the train and test set using the Word2Vec model

Now, to prepare the dataset for training and testing, first, we have to download three files, which are outlined as follows:

• A Google-trained Word2Vec model
• A large Movie Review dataset
• A sentiment labeled dataset

The pre-trained Word2Vec is downloaded from https://code.google.com/p/word2vec/ and then we can set the location for the Google News vectors manually:

public static final String WORD_VECTORS_PATH = "/Downloads/GoogleNews-vectors-negative300.bin.gz";

Then, we will download and extract the Large Movie Review dataset from the following URL.

public static final String DATA_URL = "http://ai.stanford.edu/~amaas/data/sentiment/aclImdb_v1.tar.gz";

Now, let's set the location to ...