Lexical Semantics: Reference & Sense
Referent (μ§μ λμ): λ¨μ΄λ‘ μ§μ λ μ€μ
- μ¬λ¬Όν λ¨μ΄κ° μ΄λ€ κ°λ¦¬ν€λ λμμ΄ μλ κ².
π‘ Example
Jack, the happy swimmer, my friend, and that guy can all have the same referent in the sentence Jack swims.
-> Jack = the happy swimmer = my friend = that guy
- κ°λ¨ν΄ 보μ΄μ§λ§, λκ°μ μλ―Έλ₯Ό κ°μ§κ³ μλκ±Έ μ§μΉνλ 건 μ½μ§κ° μλ€.
- μλ―Έλ₯Ό νμ ν΄μΌ μ°Ύμ μ μλ€. - μ§μ λμμ΄ κ°μλ μλ―Έκ° λ€λ₯΄λ©΄ κ°λ€κ³ ν μ μλ€.
π‘ Example
Superman, born Kal-El and legally named Clark Kent, is the last son of Krypton, sent as the dying planet's last hope to Earth, where he grew to become its kind, noble protector.
Louis Lane is in love with Superman.
Louis Lane is in love with Clark Kent??
Sense (μλ―Έ): μ°Έμ‘°μλ λ³κ°λ‘ enduringμΈ μλ―Έμ μμ
- μλ―Έ O → μ§μ체 X / μλ―Έ X → μ§μ체 O μΈ κ²½μ°λ μλ€.
π‘ Example
- μ λμ½ & νΈλΉ μ΄λΌλ λ¨μ΄λ μλ―Έλ μμ§λ§ μΈκΈμ μλ€.
- Proper names tend to have reference but no sense.
- ex. νκΈΈλ
- The same reference but different senses
- Barack Obama
- The Former President (λ―Έν©μ€κ΅μ μ μμ₯)
- Michelle Obama’s husband
Lexical Relations: Synonyms (λμμ΄)
Synonyms (λμμ΄): μΌλΆ λλ λͺ¨λ λ¬Έλ§₯μμ κ°μ μλ―Έλ₯Ό κ°μ§ λ¨μ΄λ νν.
- κ±°μ΄ λλΆλΆμ μν©μμ λμΌν μλ―Έλ‘ μ°μ΄λ κ²μ΄ λμμ΄ λΌκ³ νλ€.
- μμ ν λμμ΄λ μ΄ μΈμμ μλ€!
- λΉμ·ν΄ 보μ΄λ λ λ¨μ΄κ° λκ°μ κΈ°λ₯μ νμ§λ μλλ€.
π‘ Example
- apathetic = indifferent
- sofa = couch
- μ΄λ€ μ¬λλ€μ μ νν κ°μ μλ―Έλ₯Ό κ°μ§ λ λ¨μ΄λ μλ€κ³ μ£Όμ₯νλ€.
- λ무 vs. μΉκ΅¬
- λμ΄ vs. μ°λ Ή
- κ²½μ°° vs. μκ²½
Lexical Relations: Antonyms (λ°μμ΄)
Antonyms (λ°μμ΄): κ²ΉμΉλ κ±° μμ΄ 2κ°λ‘ λ± λλ μ§λ λ¨μ΄, λ»μ΄ μ λ°λμΈ λ¨μ΄
- Complementary[Simple] antonyms/Binary pairs
π‘ Example
- alive/dead, present/absent, awake/asleep
- alive = not dead, dead = not alive
- κ±°μ μ£½λ€, μ½κ° μ£½λ€, κ±°μ μΌμ΄λ¬λ€, μ½κ° μκ³ μλ€
- Gradable pairs[antonyms]: λΉκ΅ λμμ΄ μλ€.
π‘ big/small, hot/cold, fast/slow, happy/sad
- λ±κΈν κ°λ₯ν λ°μμ΄μ μΌλΆ μμ, νμ & νμ λμ§ μμ μ©μ΄ ν¬ν¨νλ©°, νμ λμ§ μμ μ©μ΄λ μ λ λ¬Έμ μμ μ¬μ©λλ€.
π‘ Example
How high is the mountain? not How low is the mountain?
How old are you? not How young are you?
- Relational antonyms: μλ―Έμ λμΉμ±μ λνλΈλ€.
π‘ Example: give/receive, buy/sell, employer/employee
- Antonyms(λ°μμ΄)λ νλμ μλ―Έμ νΉμ§μ μ μΈν λͺ¨λ νΉμ§μ 곡μ
π‘ κ·Έλ¬λ©΄ μ΄ λ¨μ΄λ€μ λ°μμ΄ μΈκ°μ? 1. Buy vs. sell 2. Big vs. red
Semantic Features (μλ―Έλ‘ μ νΉμ§)
- λ¨μ΄& λ¬Έμ₯μ μλ―Έμ μΌλΆμΈ κ°λ μ μμ
- Semantic Feature (μλ―Έλ‘ μ νΉμ§)μ λν Evidence
Lexical Relations (μ΄ν κ΄κ³)
Homonyms (λμ² λμμ΄μμ΄)
- μ² μκ° κ°κ³ λ°μμ΄ κ°μ λ¨μ΄
π‘ Example
- bank: μν, λ
- λ§, λ°°, μ¬κ³Ό, μν, νμ₯
Homographs (λμ² μ΄μμ΄)
- μ² μκ° κ°μ λ¨μ΄
π‘ Example
- bow: μΈμ¬νλ€, ν
- dove: λΉλκΈ°, diveμ κ³Όκ±°ν
Homophones (λμμ΄μμ΄)
- μλ―Έλ λ€λ₯΄μ§λ§ λ°μμ κ°μ λ¨μ΄
π‘ Example
- bear and bare: κ³°, λ²κ±°λ²μ
- flower and flour: κ½, λ°κ°λ£¨
Polysemy (λ€μμ΄)
- Polysemous(λ€μμ΄) λ κ°λ μ or μμ¬μ μΌλ‘ κ΄λ ¨λ μ¬λ¬ μλ―Έλ₯Ό κ°μ§ λ¨μ΄
- μλ―Έμ μΌλ‘λ κ΄λ ¨μ΄ μμ§λ§ μλ‘ λ€λ₯Έ μΈμ΄
π‘ Example
- diamond: the geometric shape; a baseball field
- κ°μ΄: κ°μ΄μ λ΄λ°λ€; κ°μ΄μ νμ°λ€
- μμΉ¨: μμΉ¨μ΄ λ°μμ¨λ€; μμΉ¨ λ¨Ήλλ€
Hyponym and hypernym (μμμ΄ & νμμ΄)
Hyponym and hypernym (μμμ΄ & νμμ΄)λ μΌλ°μ μΈ μ©μ΄μ ν΄λΉ μ©μ΄μ νΉμ μΈμ€ν΄μ€ μ¬μ΄μ κ΄κ³λ₯Ό ν¬ν¨νλ€.
rose, iris, daisy, and poppy are all a kind of flower, so rose, iris, daisy, and poppy -> κ½μ΄λΌλ λͺ¨λ λ¨μ΄μ
νμμ΄
- WordNet
https://wordnet.princeton.edu/
A large lexical database of English words. Nouns, verbs, adjectives, and adverbs are grouped into sets of cognitive synonyms called ‘synsets’, each expressing a distinct concept. Synsets are interlinked using conceptual-semantic and lexical relations such as hyponymy
π‘ Example
Argument Structure
Argument Structure (λ Όνꡬ쑰)
- μλ―Έ → λ Όνꡬ쑰 = Subcategorizationμ κ°λ€.
- μ¬κΈ°μλ λμ₯μ λμ¬ (Verb)
- Syntaxμ Constraints(μ μ½μ‘°κ±΄) λ° Subcategorization(νμ λ²μ£Όν)μ μ μ¬νλ€.
- λ€λ₯Έ μ’ λ₯μ λμ¬λ λ€λ₯Έ μμ NPλ₯Ό argumentλ‘ μ¬μ©νλ€.
- Argumentλ₯Ό κ°μ§λ©΄ μλλ€.
→ νκ΅μ΄, μμ΄λμ λ°λΌ λͺκ°μ λ Όν ꡬ쑰λ₯Ό κ°μ§λκ² λ¬λΌμ§
λ€λ₯Έ μ’ λ₯μ λμ¬λ λ€λ₯Έ μμ NPλ₯Ό μΈμλ‘ μ¬μ©νλ€.
- sleepκ³Ό κ°μ λΆμ μ¬λ λ€λ₯Έ μ£Όμ₯μ νμ§ μλλ€.
π‘ Example
- μκΈ°κ° μλ€.
- She sings. κ·Έλ λ λ Έλλ₯Ό λΆλ₯Έλ€.
- findμ κ°μ νλμ¬λ μΆκ° μΈμ(μ§μ λͺ©μ μ΄)λ₯Ό μ·¨νλ€
π‘ Example
- κ·Έλ λ¬Όμ λ§μ λ€.
- κ·Έλ λ μνμ κ°λ₯΄μΉλ€.
- giveμ κ°μ μ΄μ€λμ¬λ λ κ°μ μΆκ° μΈμ(μ§μ λͺ©μ μ΄μ κ°μ λͺ©μ μ΄)λ₯Ό μ·¨νλ€
π‘ Example
- κ½μ λ¬Όμ μ€λ€.
- Verb(λμ¬) λν λͺ¨λ μ£Όμ₯μ semantic properties (μλ―Έλ‘ μ νΉμ±)μ κ²°μ νλ€
π‘ Example
- findμ sleepκ³Ό κ°μ λμ¬λ μΈκ° μ λν μ£Όμ΄λ₯Ό νμλ‘ νλ€.
- drinkμ κ°μ λμ¬λ μ‘체 μ λν μ§μ λͺ©μ μ΄λ₯Ό νμλ‘ νλ€.
- μ² μ: human (subject)
- μ½λΌ: liquid
- κ½: plant
- λ¬Ό: liquid
- νλ³: inanimate object
- μ μ½μ‘°κ±΄ λ° νμ λ²μ£Όν (constraints and subcategorization)
- Collocational or Bag-of-words
π‘ Example
- μ½λΌ – λ§μλ€
- λ¬Ό – λ§μλ€
- κ½ – λ¬Ό
- νλ³ – λ§μλ€ ???
- μμ – λ§μλ€ ????
- κ½ – λ§μλ€ ???
κ·Έλ¬λ©΄ μ°λ¦¬λ μ΄κ²λ€μ΄ κ°μ μλ―Έλ₯Ό κ°μ§κ³ μλ€λ κ²μ μμλΌ μ μμκΉ?
- XYZ corporation bought the stock.
- They sold the stock to XYZ corporation.
- The stock was bought by XYZ corporation.
- The purchase of the stock by XYZ corporation…
- The stock purchase by XYZ corporation…
'π NLP (μμ°μ΄μ²λ¦¬) > π¨οΈ Linguistic Engineering' μΉ΄ν κ³ λ¦¬μ λ€λ₯Έ κΈ
[Words] νκ΅μ΄ ννμ & Other Morphological Processes (0) | 2024.01.16 |
---|---|
[Semantics & Pragmatics] Thematic Roles - μλ―Έμ (0) | 2024.01.16 |
[Semantics & Pragmatics] The meaning of language - μλ―Έλ‘ , μ΄μ©λ‘ (0) | 2024.01.16 |
[Syntax] Syntactic analysis in NLP - NLPμμ ꡬ문λΆμ (0) | 2023.07.26 |
[Syntax] Sentence Structure - λ¬Έμ₯ ꡬ쑰 (0) | 2023.07.26 |