fastText embeddings exploit subword information to construct word embeddings. Representations are learnt of character $n$-grams, and words represented as the sum of the $n$-gram vectors. This extends the word2vec type models with subword information. This helps the embeddings understand suffixes and prefixes. Once a word is represented using character $n$-grams, a skipgram model is trained to learn the embeddings.
Source: Enriching Word Vectors with Subword InformationPaper | Code | Results | Date | Stars |
---|
Task | Papers | Share |
---|---|---|
Text Classification | 33 | 7.91% |
General Classification | 28 | 6.71% |
Sentence | 24 | 5.76% |
Sentiment Analysis | 22 | 5.28% |
Classification | 16 | 3.84% |
Named Entity Recognition (NER) | 15 | 3.60% |
Language Modelling | 12 | 2.88% |
Word Similarity | 11 | 2.64% |
Clustering | 7 | 1.68% |
Component | Type |
|
---|---|---|
🤖 No Components Found | You can add them if they exist; e.g. Mask R-CNN uses RoIAlign |