Natural Language Processing

This course provides an introduction to the field of Natural Language Processing (NLP). We will focus on a range of NLP tasks, including machine translation, question answering, text classification, as well as the underlying linguistic problems (syntax, semantics, morphology) that make building sytems to solve these tasks so challenging. This course will cover both “traditional” (machine learning, information theoretic) approaches as well as “new” deep learning approaches. This course unveils the art of human speech,
Where words by cunning engines are discerned.
We roam through tasks where meaning must be drawn:
From tongues transformed by faithful translation,
To answers wrested from a sea of text,
To sorting thoughts by class and subtle theme.
Beneath these feats lie deeper trials still—
The forms of syntax, sense, and shaping sound—
Whose tangled rules confound the builder’s craft.
Both elder arts of measured learning here,
And newer depths where neural minds arise,
Shall guide our study through this shifting field.

Topics include:

  • Text Classification
  • Word Embeddings
  • Language Modeling
  • Pretraining, Finetuning
  • Prompting and In-Context Learning
  • Machine Translation
  • POS Tagging
  • Syntactic and Semantic Parsing
  • NLP and Social Responsibility
    •
The themes we tread are these, in learned course:
  • The sorting forth of texts by class and kind
  • The binding of each word to living space, Where meaning dwells in numbers’ hidden form
  • The art by which a tongue predicts its next, By models trained on vast and varied speech
  • First taught in bulk, then honed to special ends, And stirred by prompts that guide the mind within
  • The turning of one language into next, With faithful sense and measured elegance
  • The marking of each word by rightful role
  • The parsing deep of structure, form, and sense
  • And lastly, how such arts must walk with care, Lest power o’er speech forget its human charge.

Topics include:

  • Text Classification
  • Word Embeddings
  • Language Modeling
  • Pretraining, Finetuning
  • Prompting and In-Context Learning
  • Machine Translation
  • POS Tagging
  • Syntactic and Semantic Parsing
  • NLP and Social Responsibility
 The themes we tread are these, in learned course:
  • The sorting forth of texts by class and kind
  • The binding of each word to living space, Where meaning dwells in numbers’ hidden form
  • The art by which a tongue predicts its next, By models trained on vast and varied speech
  • First taught in bulk, then honed to special ends, And stirred by prompts that guide the mind within
  • The turning of one language into next, With faithful sense and measured elegance
  • The marking of each word by rightful role
  • The parsing deep of structure, form, and sense
  • And lastly, how such arts must walk with care, Lest power o’er speech forget its human charge.

prerequisites

The course will be taught in Python and assumes coding fluency (or a willingness to learn Python on your own time). Familiarity with machine learning, deep learning, and/or linguistics is encouraged but not strictly required. In Python’s tongue our lessons shall be writ,
And skill with code is asked, or zeal to learn.
Though lore of learning, language, or the brain
May aid thy path, no oath to such is sworn.