question archive Q6Consider the following reference translation and the output of the 4 MT systems (S1 to S4): Reference: I know it's over and it never really began
Subject:Computer SciencePrice:17.89 Bought3
Reference: I know it's over and it never really began.
S1: It's never over and it never really started.
S2: It never really began but I know it's over.
S3: It's over and it began.
S4: I know it's over.
Which system output would score higher in terms of BLEU score? Explain why.
What is the difference between generative and discriminative classification? Explain the main differences and present examples.
Describe the different levels of ambiguity that exist in human language and include at least one NLP task (e.g. POS tagging) that addresses each of them.
What is the formal definition of grammarin the context of natural language parsing systems?
Describe the main differences between word embeddings, n-gram language models, and bag of words. Provide at least one NLP application in which each of these models would be best suited for (3 applications in total, one for each model).
Purchased 3 times