Evolution of NLP using RNN, LSTM, BERT
Darshan Ramachandra Khatawakar, Sathwik Sathish, Ashwini Kodipalli, Trupthi Rao · A.A. Balkema eBooks · 2025
The field of Natural Language Processing (NLP) has seen tremendous progress, moving from the classical Recurrent Neural Networks (RNNs) to more complex models such as Long Short-Term Memory (LSTM), Robustly optimized BERT approach (RoBERTa) and Bidirectional Encoder Representations from Transformers (BERT). This development improves language generation and understanding by addressing the shortcomings of earlier models and utilizing advances in training techniques, architecture, and computer performance.