machine-learningnlpword-sense-disambiguation

Difference between fine-grained and coarse-grained score for WSD tasks?


In all Senseval and SemEval Tasks, two scores are reported - fine-grained and coarse-grained. What do they mean in the context of sense disambiguation?


Solution

  • In WordNet, you have senses which are very close. For instance, for snow you have this two senses.

    S: (n) snow, snowfall (precipitation falling from clouds in the form of ice crystals)

    S: (n) snow (a layer of snowflakes (white crystals of frozen water) covering the ground)

    In other words, while it falls from sky it is the first meaning and when it reaches the ground, it changes sense.

    In fine grained, these are two different senses while in coarse grained, it will be considered as the same sense.