Python Natural Language Processing

Para2Vec

Para2vec stands for paragraph vector. The paragraph vector is an unsupervised algorithm that uses fixed-length feature representation. It derives this feature representation from variable-length pieces of texts such as sentences, paragraphs, and documents.

Para2vec can be derived by using the neural network. Most of the aspects are the same as Word2vec. Usually, three context words are considered and fed into the neural network. The neural network then tries to predict the fourth context word. Here, we are trying to maximize the log probability and the prediction task is typically performed via a multi-class classifier. The function we use is softmax.

Please note that, here, the contexts are fixed-length and generate the context ...