Foundations of Natural Language Processing 39°

Coordinator: Andre Galassi

Room: Online via Teams and CIRSFID-ALMA AI

Objective: 10 hours of lectures + 14 autonomous practices

Ects: 3

 

LECTURES:

  • tbd - form tbd to tbd 
    Lesson 1:
    Introduction and motivation
    Basic pre-processing (tokenization, lemmatization, stemming)
    Regular Expression
  • tbd - form tbd to tbd 
    Lesson 2:
    Basic representation methods (BoW, n-grams, etc.)
    Distance matrix
    Basic tasks and methods (classification, clustering, topic modelling)
  • tbd - form tbd to tbd 
    Lesson 3:
    Advanced representation methods: word embeddings (word2vec, GloVe)
    Advanced tasks (sequence tagging, NER)
    Constituency and dependency grammars
  • tbd - form tbd to tbd 
    Lesson 4: 
    Brief introduction to recurrent architectures
    Language models
    Sequence to sequence tasks (summarization, question answering, machine translation)
  • tbd - form tbd to tbd 
    Lesson 5:
    Contextual and domain-specific embedding (ELMO, BERT, Sentence-BERT, LegalBERT etc.)
    State-of-the-art applications