MediumNLPnlp-062

Dice Bigram Coefficient

Problem

Return the Dice coefficient 2 * O / (c1 + c2) for the adjacent bigram (w1, w2), where O is how often w1 is immediately followed by w2 and c1, c2 are the token counts.

Return 0.0 when c1 + c2 is 0.

Hint

The coefficient is maximal when every occurrence of either word forms the bigram.

Test cases

case 1

in: [["new","york","new","york","city"],"new","york"]

expected: 1

case 2

in: [["the","cat","sat","on","the","mat"],"the","cat"]

expected: 0.6666666666666666

case 3

in: [["a","b","c"],"b","a"]

expected: 0

case 4

in: [[],"a","b"]

expected: 0

Starter codePython

def dice_bigram_coefficient(tokens, w1, w2):
    # Your code here
    pass

4 test cases · run your code against them in the editor