Exam: Thu 22 Oct 2026 · Lecture notes: 7 · Flashcards: 655 · Exam questions: 55
Schedule, assessment and deadlines live in the course overview.
Flashcards
655 short single-fact cards across 7 sets. Open the flashcards to study the whole course, one lecture, or any mix of lectures. Each lecture note also ends with its own set.
Exam questions: 55 long-form exam-style questions with key points and model answers, for written practice.
Lecture notes
- L01 Overview · 45 cards
What NLP is and what it models, NLP tasks (QA, NER, summarization, sentiment analysis, machine translation), NLP before and after deep learning, The role of large language models, … - L02 Multilinguality and Writing Systems · 90 cards
English-centric NLP and the resource tiers, Typological diversity across languages, Morphology, from English to Russian to Finnish, The four English-first assumptions, … - L03 Morphology and Word Formation · 92 cards
Words, vocabularies and lexicons, Word embeddings and CBOW, Vocabulary short-lists and thetoken, Word segmentation and word boundaries, … - L04 Subword Segmentation · 95 cards
The vocabulary bottleneck, Out-of-vocabulary items, Rule-based morphological analyzers (FSTs), Morfessor and Minimum Description Length, … - L05 Static Embeddings · 119 cards
Distributional semantics and count-based context vectors, Cosine similarity, Word2Vec CBOW, Word2Vec Skip-gram, … - L06 Contextual Embeddings · 102 cards
Static versus contextual word representations, Pre-training and fine-tuning workflow, Encoder-decoder, encoder-only and decoder-only architectures, BERT masked language modelling and next sentence prediction, … - L07 Crosslingual NLP · 112 cards
From multilingual to crosslingual models, Zero-shot crosslingual knowledge transfer, Parallel and comparable data, Parallel corpus acquisition (OPUS, document and sentence alignment), …
Other notes
- Overview: schedule, assessment, deadlines
- Mini Project