<과학의 목적>
1. 근본목적: 이론개발
2. 하위목적
3. 지식의 제공: 당위가 아니라 존재에 대한 관심
4. 규칙성을 알림: 이론과 법칙으로 일반화
5. 변수들 간의 관계기술: 특히 인과관계의 기술
6. 이론을 바탕으로 현상을 예측: 신뢰성 있는 예측
<과학적 연구의 특징>
1. 결정론적(모든 현
Big(“about size”) vs Red(“about color”)
: too few semantic features
- Antonym(x)
Buy(“change in possession”) vs Sell(“change in possession”)
: share all but one semantic feature(direction of the change)
- Antonym(O)
1.5.1.Semantic Features of Nouns
1.5.2.Semantic Features of Verbs
1.5.3.Semantic Features interact with different a
semantic properties of both its 1) subject
Ex. Colorless green ideas sleep furiously. -> semantically anomalous!
- “sleep” require animate subjects.
and its 2) complements.
Ex. (1) John threw/tosses/kicked/flung the boy the ball.
(2) *John pushed/pulled/lifted/hauled the boy the ball.
(3)Mary faxed/radioed/e-mailed/phoned Helen the news.
(4)*Mary murmured/mumbled/muttered/shrieked He
It seems plausible to assume that the source of the ambiguity of [an old French student] is structural in nature, and that part of the ambiguity relates to the fact that ‘old’ and ‘French’ has two different categorical functions above. We can support this analysis with empirical evidences according to syntactic, semantic, and phonological views.
1. Syntactic analysis
As you can see, [a
2.3. Compound retrieval
When using compound word, we can separate similar images such as sky and sea. This word is hard to be separated by color or shape. Text can be a useful vehicle by searching images. For example, we typed word 'blue', 'snow', and 'tree'. Although Google image retrieval is based in keyword and idee visual search lab is based in tag, we can get result which is similar to res