Designing the architecture for the model
The main blocks of the model in this example will be the following:
First, the words of the input sentence are mapped to vectors of real numbers. This step is called vector representation of words or word embedding (for more details, see Chapter 3, Encoding Word into Vector).
Afterwards, this sequence of vectors is represented by one fixed-length and real-valued vector using a bi-LSTM encoder. This vector summarizes the input sentence and contains semantic, syntactic, and/or sentimental information based on the word vectors.
Finally, this vector is passed through a softmax classifier to classify the sentence into positive, negative, or neutral.
Vector representations of words
Word embeddings are an approach to distributional semantics that represents words as vectors of real numbers. Such a representation has useful clustering properties, since the words that are semantically and syntactically related are represented by similar vectors (see Chapter 3,...