keybert

Keyword extraction technique that leverages BERT embeddings to identify the most relevant keywords in a text. KeyBERT works by using BERT embeddings to create document and candidate keyword embeddings, then calculates the cosine similarity between them to identify the most relevant keywords. It is particularly useful for tasks such as summarization, information retrieval, and improving search engine optimization (SEO).

keybert was discovered on May 2nd 2023 and it currently has a search volume of 2.9K with a growth of +2200%.