CSC 350 Comp. Ling. Exam 1 Page 1 of 4 Name _____ KEY 100 pts ________ MULTIPLE CHOICE/SHORT ANSWER. 3 points each. Choose the SINGLE BEST answer. 1. Creating a verb from a noun by adding "ize", as in computerize, is a) inflectional morphology b) derivational morphology B c) stemming d) lexical analysis 2. To parse the surface-level morphemes of words in a sentence, a language processor would normally use a) a Prolog DCG b) a context-sensitive grammar C c) an analogue of a finite automaton d) a unification-based grammar 3. A finite state transducer, when applied to a word of the lexicon would produce a a) yes/no acceptance response. b) sequence of terminals from the grammar. D c) parse tree for the word. d) sequence of constituent morphemes. 4. Lexical-level morphological parsing means finding the a) affixes of a word b) individual morphemes of a word B c) stem of a word d) spelling rules which apply to a word 5. The level at which a word is broken into its attributes such as "number" and "tense" is a) surface level b) lexical level B c) morphotactic level d) inflectional level 6. The description of semantics that explains sentences in terms of actions of a theoretical machine is called C a) denotational b) morphotactic c) operational d) axiomatic 7. Which one could be a correct translation into Prolog of the English statement "Something is worthwhile if it is good." a) worthwhile(X) :- good(Y). b) worthwhile(X) :- good(X). B c) good(X) :- worthwhile(X). d) good(X) :- worthwhile(Y). 8. Some linguists believe the human brain is structured specifically for language because a) there is definitive proof of this claim b) there is proof that animals do not have language c) there are specific brain structures which have been identified which carry out specific linguistic functions C d) every language function has been linked to a specific area of the brain 9. Early work in language translation proved more difficult than at first imagined because a) much world knowledge is required in order to translate ~any~ language. b) it was been ~proved~ impossible for any machine to ever translate any language. c) it was been ~proved~ impossible for any machine to ever translate all languages. d) certain language constructs require world knowledge in D order to be translated. CSC 350 Comp. Ling. Exam 1 Page 2 of 4 Name _____ KEY 100 pts ________ 10. Pattern recognition includes a) recognition only of visual entities. b) recognition of visual entities only by computers. c) recognition of visual and non-visual entities. C d) recognition only of non-visual entities. 11. Tagsets are used to a) define the meaning of words b) define the meaning of sentences C c) show parts of speech of words d) show parts of speech of sentences 12. Natural language processing a) seeks to endow computers with thought and feelings. b) is a completely solved problem. c) has been proven to be an unsolvable problem. D d) seeks to allow computers to understand human languages. 13. Which is not one of the three main objectives of Computational Linguistics? a) machine translation b) text retrieval D c) man-machine interfaces d) phonological awareness Consider the following Prolog code for an "action grammar". living(Verb, State) :- Verb == is, State = alive. living(Verb, State) :- Verb == was, State = dead. living(_, State) :- State = unknown. sent --> [N], [V], { living(V,S), write(N), write(' is '), write(S), nl }, [a], [Profession]. 14. Which sentence is not accepted by the grammar? a) ['Joe', was, a, mathematician] b) ['Henry', was, a, liar] c) ['Peter', hates, a, liar] d) ['Frank', is, pathological] e) neither c) nor d) is accepted D f) all are accepted by the grammar 15. Why do some rules use ':-' and some use '-->'? a) Prolog changes '-->' to ':-' and adds two parameters to each name. b) Prolog changes ':-' to '-->' and removes two parameters from each name. c) The left side of a '-->' rule is the antecedent; while the left side of a ':-' rule is the consequent. d) The left side of a '-->' rule is the consequent; while A the left side of a ':-' rule is the antecedent. CSC 350 Comp. Ling. Exam 1 Page 3 of 4 Name _____ KEY 100 pts ________ Consider the fact base: recipe(name(fish_pie),ingredients([salt(0.5,tsp), haddock(1,lb)])). recipe(name(cheese_pie),ingredients([salt(0.5,tsp), cheese(1,lb)])). recipe(name(steak_pie),ingredients([salt(0.5,tsp), steak(1,lb)])). How would Prolog respond to each? 16. | ?- recipe(name(X),_), write(X), nl, fail. fish_pie cheese_pie steak_pie no 17. | ?- X = fish_pie, recipe(name(X),Y). X = fish_pie Y = ingredients([salt(0.5,tsp), haddock(1,lb)]) yes Consider the following context free grammar, G, expressed as a Prolog DCG, for a language of noun phrases only. snt --> det, adjs, noun. det --> [a] ; [the] ; []. adjs --> adj ; adj, adjs. adj --> [spotted] ; [endangered] ; [poison_arrow]. noun --> [frog] ; [owl]. 18. Produce a complete leftmost derivation of the sentence: the spotted endangered owl (Note: in Prolog it would have to be in the form [the, spotted, endangered, owl]) snt => det, adjs, noun => [the] adjs noun => [the] adj adjs noun => [the] [spotted] adjs noun => [the] [spotted] adj noun => [the] [spotted] [endangered] noun => [the] [spotted] [endangered] [owl] 19. If you entered snt(X,[]) at the Prolog prompt, what would be the first sentence that Prolog generated? X = [a, spotted, frog] 20. To accommodate surface feature agreement, some rules are changed to: snt --> det(F), adjs(F), noun. det(F) --> [a], { F = consonant } ; [an], { F = vowel } ; ... Which is the correct way to change adjs? a) adjs(Feature) --> adj(Feature); adj(Feature), adjs(Feature). b) adjs(Feature) --> adj(Feature); adj(Feature), adjs(_). c) adjs(Feature) --> adj(Feature); adj(Feature), adjs. B d) adjs(Feature) --> adj; adj, adjs(Feature). CSC 350 Comp. Ling. Exam 1 Page 4 of 4 Name _____ KEY 100 pts ________ MATCHING. Match each term with its BEST definition. No choice will be used more than once; some will not be used. 2 points each. ____ ---- a,b | ---- | a,b ---- _______ _w_ 21. | |--------->|| ||--------->| | | b | s0 | || s1 || | s2 |<------ ---- | ---- |<--------- ---- ---- a ____ ---- a,b | ---- | b ---- _______ _x_ 22. | |--------->|| ||--------->| | | a,b | s0 | || s1 || | s2 |<------ ---- | ---- | ---- ---- | ^ |_a_| _e_ 23. How the sentences of a language are formed. _t_ 24. Construction of a form of a word's root. _f_ 25. The meanings of the sentences in a language. _k_ 26. Inference required for a sentence to make sense. _a_ 27. How sounds are used in language. _r_ 28. Change in a word's form to reflect a change in its part of speech. _g_ 29. The study of the constraints in word formation. _n_ 30. Automatic inference of meaning from natural language texts. _c_ 31. Mastery of the underlying rules of a language. _d_ 32. Application of the underlying rules of a language. _l_ 33. Meaning not directly derived from a component of a sentence. _m_ 34. The distinct meaning of a sentence in context. _h_ 35. A word category that is provided with mechanisms for expansion. _p_ 36. A mediator between humans and machines. _s_ 37. Recognition of human language. _u_ 38. Morphological addition to an existing word. _b_ 39. How words are formed in language. _z_ 40. {a, b} {0, 1}* {c, d} {e, f}* a) phonology n) information retrieval b) morphology o) machine translation c) competence p) man-machine interface d) performance q) inflection e) syntax r) derivation f) semantics s) natural language processing g) morphotactics t) stemming h) open class u) affix i) closed class v) finite-state transducer j) cancellation w) machine accepting (a+b)((a+b)b*a)* k) presupposition x) machine accepting (a+b)a* l) implicature y) regular expression matching abc m) pragmatics z) regular expression matching a0c