GloVe and fastText — Two Popular Word Vector Models in NLP

Miklov et al. introduced the world to the power of word vectors by showing two main methods: Skip–Gram and Continuous Bag of Words (CBOW). Soon after, two more popular word embedding methods built on these methods were discovered. In this post, we’ll talk about GloVe and fastText, which are extremely popular word vector models in the NLP world.

Global Vectors (GloVe)

Pennington et al. argue that the online scanning approach used by word2vec is suboptimal since it does not fully exploit the global statistical information regarding word co-occurrences.

How NLP Is Teaching Computers the Meaning of Words

Humans are good at conversations. We understand what someone means when they say something and can understand when a word like “bank” is used in the context of a financial institute or a river bank. We use the power of logical, linguistic, emotional reasoning and understanding in order to respond during conversations.

In order to get machines to truly understand natural language like we do, our first task is to teach them the meaning of words — a task that is easier said than done. The past couple of years has seen significant progress in this field.