I am trying to understand what is word-level embedding. From what I found, it is a vector representation of a word which captures the context information in which the word is used.

I don't understand how this is represented in a vector and no simple example on-line at all.

Can anyone explain or demonstrate how it works?

Thank you


Word embedding means how words embedded into learning algorithms. For example, "previous word feature", which means the previous word of current word, with a value of "hello", we should transform "hello" of string form into numerical vectors so that machine learning algorithms such as logistic regression and SVM can understand what you mean (They do not understand string "hello").

To obtain word embeddings, there are several different ways. I guess methods where "context information in which the word is used" are something like Mikolov's Skip-gram model or CBOW, or Bengio's Neural network language model.

Actually, not only these models can train word representations (or word embeddings), traditional ways like Vector space model, Bag-of-word model, one-hot coding, and PCA, can also provide word embeddings.

| cite | improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.