Scope ambiguity in natural language

Question

I feel it is bit curious to understand the Natural language processing. I have the following questions..

What is meant by Scope ambiguity in natural language?

How can done Statistical resolution of scope ambiguity?

Which is the best language can I use for the Statistical resolution?

Answer 1

see:

www.sfu.ca/~jeffpell/Ling324/fjpSlides7.pdf

nlp.stanford.edu/projects/nlkr/scoper.pdf

Answer 2

Scope ambiguity refers to the order of precedence of quantifiers (words like "a", "the", "each", "some", "every", "all", "one" etc.) in a natural language sentence.

For example, consider this sentence: "The dog brings me the newspaper every morning".

You know that the sentence parses as: "Exists DOG d ( Foreach MORNING m ( Exists NEWSPAPER n ( d brings n during m ) ) )". In other words, every morning, the newspaper is different.

But a computer program might instead interpret the sentence to mean "Exists DOG d ( Exists NEWSPAPER n ( Foreach MORNING m ( d brings n during m ) ) )" - in other words, there is one old newspaper that you haven't thrown away, and every morning, the dog brings it to you.

Resolving scope ambiguity, as far as I know, is very much an unsolved problem.