How to study Natural Language Processing — Electives, Pokhara University BECE
What it covers
Teaching computers to work with human language: text preprocessing, text representation, language models, text classification, word embeddings, and transformer models.
Topics worth extra time
- Text preprocessing: tokenisation, stemming and lemmatisation
- Language models and n-grams
- Part-of-speech tagging and parsing
- Word embeddings and text classification
- Neural NLP and transformers
How to study it
Use Python libraries such as NLTK or spaCy on real text, then try a pretrained transformer model. Work n-gram probabilities by hand on small corpora.
In the exam
Show probability calculations clearly and explain each pipeline stage with an example sentence.