MediumNLPnlp-162

Slot BIO Decode

Problem

Decode slot tags into a dictionary mapping each slot label to the list of extracted entity strings.

B- starts a new value, I- continues the current slot when the label matches, an I- with a different label starts a new value of that label, and O closes the current slot. Multi-token values are joined with spaces.

Hint

Flush the current value whenever the label changes or an O appears.

Test cases

case 1

in: [["fly","to","paris"],["O","O","B-city"]]

expected: {"city":["paris"]}

case 2

in: [["new","york","to","boston"],["B-city","I-city","O","B-city"]]

expected: {"city":["new york","boston"]}

case 3

in: [["i","am","here"],["O","B-state","O"]]

expected: {"state":["am"]}

case 4

in: [[],[]]

expected: {}

case 5

in: [["a","b"],["B-x","I-y"]]

expected: {"x":["a"],"y":["b"]}

Starter codePython

def slot_bio_decode(tokens, tags):
    # Your code here
    pass

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