Linguistics for Natural Language Processing
The linguistic foundations of NLP, AI, and language understanding
Language Looks Simple. It Isn't.
Click each sentence to reveal its hidden ambiguity — and what that means for any NLP system trying to process it.
“I went to the bank.”
A financial institution — I need to deposit money or see a cashier.
The edge of a river — I walked down to the water's edge.
“I saw the man with the telescope.”
I used a telescope to observe him.
The man I saw was carrying a telescope.
“The police stopped the protesters because they were becoming violent.”
The protesters were becoming violent, so the police intervened.
The police were becoming violent, and they stopped the protesters.
“Every student speaks two languages.”
There are two specific languages — say English and French — that every student speaks.
Each student speaks some two languages — not necessarily the same ones as other students.
“Visiting relatives can be boring.”
Relatives who come to visit — those people — can be boring.
The activity of going to visit relatives can be boring.
0 of 5 revealed
Course Structure
Ten units building from word structure to sentence structure, meaning, and AI applications. Based on Bender (2013), Linguistic Fundamentals for Natural Language Processing.
From what language is, to morphemes, phonological alternations, and grammatical feature systems. Concepts #0–#43.
Syntax, parts of speech, heads, arguments, adjuncts, and grammatical functions. Concepts #44–#82.
When linguistic form and meaning diverge — passives, expletives, long-distance dependencies, argument drop. Concepts #83–#97.
Morphological analysers, deep parsers, typological databases, and the future of linguistically informed AI. Concepts #98–#100.