How to take the weighted average of fast-text embedding using TF-IDF as weights of each word

148 Views Asked by At

If I have textual data and I want to convert it to word embedding using fast-text but instead of taking the average of each vector to represent the sentence, how can I take the weights of the same text of TF-IDF per word level and multiply each word(as word embedding) with its weight acquired from TF-IDF.

0

There are 0 best solutions below