Embeddings

5.1 Lexical semantics

1. Why is representing a word’s meaning as DOG or LIFE’ unsatisfactory?
   It says nothing about meaning: it can’t tell us that cat is similar to dog, cold is the opposite of hot, happy is positive, or that buy, sell and pay describe one event from different perspectives.
   • Jurafsky & Martin, Speech and Language Processing, 3rd ed. draft, ch. 5, §5.1, p. 128
2. What is lexical semantics?
   The linguistic study of word meaning.
   • Jurafsky & Martin, Speech and Language Processing, 3rd ed. draft, ch. 5, §5.1, p. 128
3. What is the difference between a lemma and a wordform?
   The lemma (citation form) is the dictionary form: mouse for mice, sing for sang, dormir for Spanish duermes. Wordforms are the specific inflected forms that occur.
   • Jurafsky & Martin, Speech and Language Processing, 3rd ed. draft, ch. 5, §5.1, p. 128
4. What is a word sense, and what does it mean for a lemma to be polysemous?
   A sense is one aspect of a word’s meaning (mouse the rodent vs the cursor device). A polysemous lemma has multiple senses; choosing the right one in context is word sense disambiguation.
   • Jurafsky & Martin, Speech and Language Processing, 3rd ed. draft, ch. 5, §5.1, p. 128
5. Define synonymy, informally and formally.
   Two senses are synonyms if their meanings are (nearly) identical (couch/sofa, car/automobile). Formally, two words are synonymous if they are substitutable in any sentence without changing its truth conditions.
   • Jurafsky & Martin, Speech and Language Processing, 3rd ed. draft, ch. 5, §5.1, p. 129
6. What is the principle of contrast, and what does it imply for synonymy?
   Any difference in linguistic form comes with some difference in meaning (e.g. H₂O is scientific, water fits a hiking guide). So true synonymy hardly exists; “synonym” really means approximate synonymy.
   • Jurafsky & Martin, Speech and Language Processing, 3rd ed. draft, ch. 5, §5.1, p. 129
7. How does word similarity differ from synonymy?
   Few words have synonyms, but most have many similar words (cat and dog). Similarity is a graded relation between words rather than between senses, which avoids committing to a representation of senses.
   • Jurafsky & Martin, Speech and Language Processing, 3rd ed. draft, ch. 5, §5.1, p. 129
8. How can we get human judgments of word similarity? Name a dataset.
   Ask people to rate how similar word pairs are. SimLex-999 rates pairs on a 0–10 scale, from near-synonyms (vanish/disappear) to unrelated pairs (hole/agreement).
   • Jurafsky & Martin, Speech and Language Processing, 3rd ed. draft, ch. 5, §5.1, p. 129
9. What is word relatedness (association), and how does it differ from similarity?
   Words can be related without being similar: coffee and cup share no features but co-participate in an everyday event; scalpel and surgeon are related eventively.
   • Jurafsky & Martin, Speech and Language Processing, 3rd ed. draft, ch. 5, §5.1, p. 129
10. What is a semantic field?
    A set of words covering a semantic domain and bearing structured relations to each other, e.g. hospitals (surgeon, scalpel, nurse, anesthetic) or restaurants (waiter, menu, plate, chef).
    • Jurafsky & Martin, Speech and Language Processing, 3rd ed. draft, ch. 5, §5.1, p. 129
11. What are topic models?
    Unsupervised methods like Latent Dirichlet Allocation (LDA) that induce sets of associated words (akin to semantic fields) from large text collections, useful for discovering topical structure in documents.
    • Jurafsky & Martin, Speech and Language Processing, 3rd ed. draft, ch. 5, §5.1, p. 130
12. What is connotation in the sense used here?
    The aspects of a word’s meaning related to emotions, sentiment, opinions or evaluations: fake/knockoff/forgery vs copy/replica/reproduction, innocent vs naive.
    • Jurafsky & Martin, Speech and Language Processing, 3rd ed. draft, ch. 5, §5.1, p. 130
13. What are Osgood’s three dimensions of affective meaning?
    Valence (pleasantness), arousal (intensity of emotion provoked) and dominance (degree of control exerted). Happy is high valence, excited high arousal, controlling high dominance.
    • Jurafsky & Martin, Speech and Language Processing, 3rd ed. draft, ch. 5, §5.1, p. 130
14. Why was Osgood’s affective-meaning model revolutionary for vector semantics?
    Representing each word by three numbers made a word a point in a three-dimensional space: the first expression of the idea that meaning can be a vector.
    • Jurafsky & Martin, Speech and Language Processing, 3rd ed. draft, ch. 5, §5.1, p. 130