Introduction to CL Session 1: 7/08/2011 What is computational - - PowerPoint PPT Presentation
Introduction to CL Session 1: 7/08/2011 What is computational linguistics? Processing natural language text by computers for practical applications ... or linguistic research Among practical applications Sometimes the
for practical applications
... or linguistic research
Sometimes the computer only needs to classify or transform
the text
... but sometimes it needs to “understand”
Ex: Watson: winner of ‘Jeopardy’ CL vs. NLP (natural language processing)
– He visited New York in 2003.
– visited visit + -ed
– He/Pron visited/V New/?? York/N in/Prep 2003/CD
– He visited [LOCATION New York] in [YEAR 2003]
– [NP He] [V visited] [NP New York] in [NP 2003]
– (S (NP (PRON he)) (VP (V visited) ….)
– Semantic tagging: *AGENT He+ visited *DEST New York+ …. – Meaning: visit (he, New-York)
– Co-reference: “He” refers to “John” – Discourse structure
– “too”, “two”, “to” – “ice cream” vs. “I scream” – “ta” in Mandarin: he, she, or it
– unlockable: [[un-lock]-able] vs. [un-[lock-able]]
– John saw a man with a telescope. – Time flies like an arrow.
– Ex: “bank”, “saw”, “run”
– Ex: every boy loves his mother – Ex: John and Mary bought a house
– Susan called Mary. She was sick. (coreference resolution) – It is pretty hot here. (intention resolution)
– “brother”, “cousin”, “uncle”, etc.
– Parsing:
– WSD:
– MT:
– Require training data – Build a statistical model – Knowledge and rules can be incorporated into the model as features etc.
– Monolingual: English/Chinese/Arabic Gigawords – Parallel data: UN data, EuroParl
– Syntactic treebanks: a set of parse trees – Proposition Bank: – Discourse Treebank
Applications Task1 Task2 Task_i ML1 ML_m ML2 … D1 D2 D_n … …
– Good: it can be ported to many languages without any changes. – Bad: it cannot take advantage of properties of certain languages.
– the design of models – as features – as filters – … Building a treebank is an effective way.