WordNet

• Description: WordNet is an on-line lexical reference system in which English nouns, verbs, adjectives and adverbs are organized into synonym sets, each representing one underlying lexical concept. Different relations link the synonym sets.
  WordNet was developed in the Cognitive Science Laboratory at Princeton University under the direction of Professor George Miller.
• Availability:

  WWW Interface

  http://www.cogsci.princeton.edu/~wn/w3wn.html

  Source Distributions

  Unix (9.1MB), PC (5.8MB), Macintosh (7.5MB), Prolog (database only, 4.2MB). ftp://clarity.princeton.edu/pub/wordnet/

  Extended interfaces developed by WordNet users (for X, Lisp etc) are listed in the WordNet home page.
• Further information: Email: wordnet@princeton.edu
  WWW: WordNet home page: http://www.cogsci.princeton.edu/~wn/
  README: ftp://clarity.princeton.edu/pub/wordnet/README
  Publications: ftp://clarity.princeton.edu/pub/wordnet/5papers.ps

Lexical relations

• synonyms (e.g. hate and detest) or
• antonyms (such as push and pull).

• cat is a hyponym of feline.
• and feline is a hypernym of cat!
• meronym relationship (part-whole relation), for example, table has leg and tabletop as meronyms.

All of the WordNet noun synsets are organized into hierarchies. Each synset is part of at least one hierarchy, headed by a synset called a unique beginner. All of these synsets originate in the lexicographer file noun.Tops. From any noun synset (excluding the unique beginners), the hypernym pointers can be traced up to one of the following unique beginners:

{ entity, (something having concrete existence; living or nonliving) } { psychological_feature, (a feature of the mental life of a living organism) } { abstraction, (a concept formed by extracting common features from examples) } { location, space,#p (a point or extent in space) } { shape, form, (the spatial arrangement of something as distinct from its substance) } { state, (the way something is with respect to its main attributes; "the current state of knowledge"; "his state of health"; "in a weak financial state") } { event, (something that happens at a given place and time) } { act, human_action, human_activity, (something that people do or cause to happen) } { group, grouping, (any number of entities (members) considered as a unit) } { possession, (anything owned or possessed) } { phenomenon, (any state or process known through the senses rather than by intuition or reasoning) }

Synsets: example from VIBRATION

• -> motion -- a state of change; "they were in a state of steady motion".
  
  • => shaking, tremor, trembling, quiver, quivering, vibration, palpitation
    
• => precession -- the motion of a spinning body (as a top) in which it wobbles so that the axis of rotation sweeps out a cone.
  
  • => stream, flow -- something that resembles a flowing stream in moving continuously; "a stream of men".

Database organization

COMLEX Syntax

• a monolingual English Dictionary consisting of 38,000 head words intended for use in natural language processing.
• developed by the Proteus Project at New York University under the auspices of the Linguistic Data Consortium (LDC)
• includes entries for approximately 21,000 nouns, 8,000 adjectives and 6,000 verbs
• Nouns have 9 possible features and 9 possible complements;
• adjectives have 7 features and 14 complements
• verbs have 5 features and 92 complements
• For 750 frequent verbs, there are an additional 4 possible features and 32 possible complements.

(verb           :orth "build" 
                :subc ((np) (np-for-np) (part-np :adval ("up")))) 
(noun           :orth "assertion" 
                :subc ((noun-that-s) (noun-be-that-s))) 
(adverb         :orth "even") 
(adjective      :orth "above-mentioned" 
                :features ((apreq) (attributive))) 
(verb           :orth "abbreviate" 
                :subc ((np-pp :pval ("to")) (np) (np-np-pred) (np-as-np))
                :features ((vveryving :pastpart t))) 
(noun           :orth "Prof." 
                :features ((ntitle))) 

a lexicon of English semantics based on Frame Semantic principles

• FrameNet is a collaboration between linguists, lexicographers, and computer scientists, (and other NLP researchers in the US and abroad) to fill a significant gap in the types of machine-readable lexicons currently available: a detailed semantic description of a substantial portion of the English lexicon, coded for computational use, and annotated with references to corpus examples.
• Intends to meet both the breadth and depth requirements for NLP uses
• Database contents:
  • explicit information about word senses,
  • the syntactic and semantic combinatory possibilities of many words (e.g. syntactic and semantic "subcategorization" and "predicate-argument" information),
  • the kinds of semantic frames into which they can enter,
  • and the probabilities of the different senses and frame combinations of these words.
• enhancements to WordNet entries
• 5,000 lexical units representative of the semantic domains of health care, chance, perception, communication, transaction, time, space, body, motion, stages of life, social context, emotion, and cognition.

Part of Frame-semantic "Tagset" for the Health Frame

KWIC for some corpus sentences for the lemma CURE

Frame Semantic Enhancements of WordNet Sentence Frames

A subframe can inherit elements and semantics from its parent.

frame (CommercialTransaction)
frame-elements\{ buyer,  seller,  payment,  goods\
scenes ( buyer gets  goods, 
 seller gets  payment)

frame (RealEstateTransaction)
inherits (CommercialTransaction)
link( borrower =  buyer,  loan =  
payment)
frame-elements\{ borrower,  loan,  lender\
scenes ( loan (from  lender) creates  payment,
       buyer gets  loan)
A subframe can inherit elements and semantics from its parent.