CSC 350 Comp. Ling. Exam 3 Page 1 of 6 Name _____ KEY 160 pts ________ Multiple Choice/Short Answer. 3 points each. Choose the SINGLE BEST answer. 1. 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 2. Attributes such as "number" and "tense" would be regarded as a) surface level b) lexical level B c) intermediate level d) inflectional level 3. Creating a verb from a noun by adding "ize", as in computerize, is a) stemming b) lexical analysis D c) inflectional morphology d) derivational morphology 4. Natural language processing is a) mapping between two or more human languages. b) text retrieval from a data warehouse based on human language input. c) automated extraction of meaning from human language. C d) conversion of spoken words to symbols. 5. The parse tree of a sentence is a) inherent in the language b) inherent in the grammar B c) identical for all grammars d) identical for all sentences 6. Some grammars are arguably "better" than others because a) their associated parse trees correspond to the language's syntax. b) their associated parse trees correspond to the language's semantics. c) their associated derivations are top-down. B d) their associated derivations are bottom-up. 7. Tagsets are used to a) define the meaning of words b) show parts of speech of words B c) add attributes to grammars d) add actions to CFG's 8. Which describes an aspect of the Chomsky hierarchy of languages? a) Type 0 languages are equivalent to Turing machines. b) Type II languages are equivalent to finite automata. c) Type III languages are the Context Free Languages. A d) Type 0 and Type III languages are exactly the same. 9. The X-Bar hypothesis a) has been shown to be the mechanism the brain uses to decode sentences. b) is provably better than any programmable alternatives. c) places specifiers to the right of nouns in all languages. D d) organizes linguistic elements as binary trees. 10. In Interlingual Translation, the source language is translated a) into a language-independent representation before being translated into the target language. b) into a language-independent representation after being translated into the target language. c) directly into the target language. A d) directly into phonological output. CSC 350 Comp. Ling. Exam 3 Page 2 of 6 Name _____ KEY 160 pts ________ 11. Which regular expression describes the language of the given machine? A a) (v+w)v* b) (v+w)(v*(ww*v)*)* c) v + w + ww c) v + w + w(v+w) ____ ---- v,w | ---- | w ---- _______ -->| q |--------->|| q ||--------->| q | | v,w | 0 | || 1 || | 2 |<------ ---- | ---- | ---- ---- | ^ |_v_| 12. Representational ambiguity a) is automatically solved by using predicate logic as the representation. b) can be resolved by using canonical forms for knowledge representation. c) is caused by vagueness in natural language. B d) is caused by using an ambiguous grammar for the language. 13. A semantic attachment specifies a) how to compute the meaning representation of an entity from the meanings of its constituents. b) whether to use denotational, axiomatic, or operational semantics. c) how a unification grammar attaches features to on element of the parse tree. A d) the set of tags attached to a corpus. 14. Lexical semantics a) relies entirely on a lexicon to ascribe meaning to a sentence. b) relies entirely on a grammar to ascribe meaning to a sentence. c) ascribes much of the meaning of a sentence to individual words. C d) is based strictly on part-of-speech tagging. 15. A sentential form is any string of a) productions. b) terminals. c) terminals and nonterminals. D d) terminals and nonterminals derivable from the start symbol. 16. Template-based approaches to speech recognition a) require visual methods to match patterns in the templates. b) compare input speech to stored units. c) analyze input speech statistically to match patterns. B d) have had no success in the translation process. 17. The entity for which a symbol stands is called a(n) C a) cataphor b) anaphor c) referent d) subject 18. Referring to a previously introduced symbol is called C a) coreference b) situational context c) anaphora d) cataphora 19. Human dialogue is characterized by A a) turn-taking b) role-playing c) information exchange d) illocution 20. VoiceXML is meant a) for Web-based dialogue control. b) purely to control Internet telephony. c) for natural language processing. A d) only for turn-taking dialogues. CSC 350 Comp. Ling. Exam 3 Page 3 of 6 Name _____ KEY 160 pts ________ 21. Phonology has to do with how a) individual letters are sounded b) pieces of words are sounded B c) complete words are sounded d) complete sentences are sounded 22. Inflection has to do with a) tense and number b) derivation (etymology) A c) sound (pitch, duration) variations d) length of words 23. Explaining the meaning of a Java program by describing how its statements change the state of the Java Virtual Machine is an example of a) denotational semantics b) axiomatic semantics C c) operational semantics d) analogical semantics 24. Syntax can be captured in B a) a vocabulary b) a grammar c) a set of phonemes d) an alphabet 25. Different lexemes sharing one meaning is C a) polysemy b) homonymy c) synonymy d) antonymy 26. Garden path sentences arise because a) a sentence has more than one meaning. b) the meaning representation of a sentence is ambiguous. c) an incorrect parse tree is initially constructed. d) the speaker intends to mislead. 27. Semantics a) can be captured as the sum of all component parts of a sentence b) can be captured in a grammar c) depends on constituent elements and their positions within a sentence C d) is the mapping between surface and lexical level elements of a sentence 28. When performing translation using templates, simplification rules help to a) remove inconsistencies from the meanings of statements. b) make the source language smaller. c) make the source language larger. d) map source language statements with the same semantics D to the same target language statement. 29. Which is not one of the structures of Discourse Structure? D a) linguistic b) intentional c) attentional d) intrinsic 30. Presupposition: what is presupposed by the sentence? The president of this college is every bit as capable as any man holding that office would be. a) that the president is a woman b) that the president is capable A c) that the president is incapable d) that the college has no president 31. Cancellation: what is cancelled in this sentence? The gubernatorial candidate was the inarticulate son of a former politician, in the movie "Silver City." a) that the candidate was running for governor of Texas b) that Silver City is a real place c) that the candidate for governor was inarticulate C d) that the movie was about a real person CSC 350 Comp. Ling. Exam 3 Page 4 of 6 Name _____ KEY 160 pts ________ 32. Which did not originate in France? a) systematization of sign language b) the Statue of Liberty D c) Prolog d) Jerry Lewis 33. That a meaning representation scheme should possess verifiability means a) a sentence should have only one representation. b) there should be only one way to model the world at any one time. c) predicate logic can be used to prove that the representation models the world correctly. d) there must be a defined relationship between the representation and D the real world. 34. Semantic ambiguity a) is an unavoidable consequence of natural language. b) is caused by representational ambiguity. c) can be resolved by using canonical forms for knowledge representation. A d) can be resolved by using an unambiguous grammar for the language. 35. Window is a meronym of B a) portal b) house c) view d) pane 36. Which is the best example of how anaphoric usage can cause unintended interpretation of meaning? a) John hit the man with his book. b) John put the food in his mouth. c) I saw the man with the telescope. A d) The rotten man chased the dog. 37. With the given grammar, in what order will Prolog supply sentences? sentence --> np, vp. np --> adj, noun. vp --> verb, np. adj --> [deranged]. noun --> [terrorist]. noun --> [nuke]. verb --> [steals]. ?- sentence(X,[]). a) X = [deranged, terrorist, steals, deranged, terrorist] ; X = [deranged, terrorist, steals, deranged, nuke] ; X = [deranged, nuke, steals, deranged, terrorist] ; X = [deranged, nuke, steals, deranged, nuke] b) X = [deranged, terrorist, steals, deranged, terrorist] ; X = [deranged, nuke, steals, deranged, terrorist] ; X = [deranged, terrorist, steals, deranged, nuke] ; X = [deranged, nuke, steals, deranged, nuke] c) X = [deranged, nuke, steals, deranged, nuke] ; X = [deranged, nuke, steals, deranged, terrorist] ; X = [deranged, terrorist, steals, deranged, nuke] ; X = [deranged, terrorist, steals, deranged, terrorist] d) X = [deranged, terrorist, steals, deranged, terrorist] ; A no CSC 350 Comp. Ling. Exam 3 Page 5 of 6 Name _____ KEY 160 pts ________ 38. A bottom-up parse of the sentence "a dog barks" is shown. What rule can you infer is part of the underlying grammar based on the line labeled Z? _ - a dog barks a - dog barks D - dog barks D dog - barks D N - barks <-----Z NP - barks NP barks - NP V - NP VP - S - a) NP --> D N b) N --> dog c) D --> a B d) D N --> NP Assume a knowledge base like the following. (The entire export/3 fact base is not necessarily shown - this is just a sample.) % export(Country,Product,Specific_Type). export('Great Britain',cheese,'Penbryn'). export('Greece',cheese,'Kaseri'). export('Denmark',cheese,'Saga'). export('United States',oil,corn). export('Greece',oil,olive). export('Spain',oil,soybean). export('Canada',oil,canola). 39. Write a Prolog grammar rule to recognize the following sentence, Who exports Specific_Type ? where Specific_Type is any word at all. For example, Who exports Penbryn ? sentence --> ['Who'], [exports], [Product], ['?']. 40. Add a Prolog action to the rule you created in the previous problem to respond appropriately. Be sure to handle the case where Specific_Type isn't in your database. Assume at most 1 country exports the specific type of product. your-lhs --> your-rhs, { ACTION }. Answer: sentence --> ['Who'], [exports], [Type], ['?'], { export(Who,_,Type), write_ln(Who); write_ln('I don''t know.') }. CSC 350 Comp. Ling. Exam 3 Page 6 of 6 Name _____ KEY 160 pts ________ True or False. 2 points each. _T_ 41. abc is in the language = {a, b}* {0, 1}* {c, d} {e, f}* _F_ 42. a01c is in the language = {a, b} {0, 1} {c, d} {e, f}* _T_ 43. In Machine Translation, direct translation does not require elaborate analyses of the underlying structures of a sentence. _F_ 44. A corpus contains mappings to the morphemes of its words. _T_ 45. There are more phonemes than "visemes" in the English language. _F_ 46. The statement "She's the sharpest woman to ever graduate from our program, presupposes that she is intelligent. _F_ 47. Illocutionary acts are those spoken for their meaning. _T_ 48. Rhetorical structure refers to the relationships among elements of the discourse. _T_ 49. In pronoun resolution, a pronoun used as the subject would have a higher salience than one used as an indirect object. _F_ 50. In discourse model update, an entity's salience is increased as each new sentence is encountered. _T_ 51. In the discourse structure model attentional state refers to methods of modeling where the focuses of dialogue participants lie. _T_ 52. In the example "John has a new Acura. It is red.", the intersentence co-referring expressions are "Acura" and "it". _T_ 53. A grammar with extra elements bound to its tokens is called a feature grammar. _F_ 54. A sequence of production applications which always substitutes for the left-most nonterminal is called bottom-up parsing. _T_ 55. Substitutability is the ability to replace a word with its synonym without altering the meaning of the sentence. _F_ 56. The goal of "information retrieval" as related to NLP is mostly met by stemming. _T_ 57. The goal of "machine translation" is to automatically translate one human language into another. _F_ 58. The term "open" used to describe a word class (like noun) means these words can be used for other parts of speech (like a verb). _T_ 59. Adding "es" to a word that ends in "ss", like "mess" or "fuss", to form the plural is an example of surface-level morphology. _T_ 60. Determining that a word ending in "es", like "messes", is a plural instead of a present tense verb is an example of lexical-level morphology.