COS4861 2030/2020 Natural Language Processing Year module

COS4861 203/0/2020 Natural Language Processing Year module/ COS4861 203/0/2020 Natural Language Processing Year module ASSIGNMENT 3 Solution Total Marks: 50 Unique Assignment Number: 886028 Please note that it is your responsibility to check that your assignment is registered on the assignment database, whether you submitted the assignment by post or via myUnisa. You can do this by visiting myUnisa. Question 1: 10 Marks Provide a formal definition of a PCFG. Explain how PCFGs can aid in disambiguation, provide an example to aid your discussion. Formal definition is in section 14.1  N A set of non-terminals   a set of terminals distinct from N  R a set of productions of the form A [p] where 1.A 2 N, 2.  2 ( [ N)  , 3. 0  p  1, with p = P(jA)  S is a designated start symbol. 5 marks Discussion on how PCFGs can be used in disambiguation section 14.1.1 5 marks. Expected: a general discussion on how the probabilities can be used to calculate which parse tree of twee ambiguous parses of the same sentence can be used based on the probabilities calculated using the PCFG. An example to accompany this. f0g Question 2: 5 Marks Using the online dictionary WordNet, give a detailed account of similarities and differences among the following set of lexemes: simulated, synthetic and artificial. Simulated is different from synthetic: synthetic refers to man made, or false, where simulated could mean to model, and actual event, or to provide a close approximation of something. They are similar where both refer to a falsity of the object under consideration. Simulated is different from artificial: where artificial refers to something is unreal (contrived by art rather than nature), or simply false. It is similar to artificial, where they both refer to made made models or events.