NLP Study Plan

What is NLP, history of natural language processing, core tasks, text as data, challenges in language understanding, and real-world applications.

Tokenization techniques, stemming and lemmatization, stopword removal, regular expressions for text, normalization, and text encoding strategies.

Bag of words model, TF-IDF weighting, N-gram language models, one-hot encoding, co-occurrence matrices, and dimensionality reduction techniques.

Word2Vec CBOW and Skip-gram architectures, GloVe embeddings, FastText subword representations, embedding evaluation, and subword embeddings.

Recurrent neural networks, LSTMs for long-range dependencies, GRUs, bidirectional RNNs, sequence-to-sequence models, and encoder-decoder architectures.

Attention basics and intuition, self-attention, multi-head attention, positional encoding, the Transformer architecture, and attention variants.

Sentiment analysis, document classification, multi-label classification, CNNs for text, fine-tuning pretrained models, and evaluation metrics.

NER approaches and architectures, BIO tagging scheme, CRF layers for sequence labeling, SpaCy NER, evaluation metrics, and custom NER training.

Statistical language models, neural language models, perplexity metric, BERT architecture, GPT architecture, and masked vs causal language modeling.

Parallel corpora, word alignment, sequence-to-sequence with attention, Transformer-based MT, BLEU score evaluation, and multilingual models.

Extractive question answering, generative QA, passage retrieval, dense retrieval methods, reading comprehension, and open-domain QA systems.

Extractive summarization, abstractive summarization, beam search decoding, nucleus sampling, controlling generation, and evaluation metrics.

Zero-shot learning, few-shot learning, prompt engineering, multimodal NLP, ethics and bias in NLP, and research frontiers.