BERT (Bidirectional Encoder Representations from Transformers) is a recent method to emerge from the groundbreaking R&D happening within deep learning. BERT is changing the game in natural language processing. Some of the profound benefits BERT brings to AI include:
- Much better model performance over legacy methods
- An ability to process larger amounts of text and language
- An easy route to using pre-trained models (transfer learning)
- Capabilities to fine tune your data to the specific language context and problem you face
For certain situations, BERT can even be applied directly to the data with no further training (in other words, zero-shot training) and still deliver a high-performing model.