Tagging: Problems with the dash?

Remember that we are giving some kind of special treatment to the dash? This had to be done in order to pick up things like ‘a-line’. Here is another example where the description contains a dash, but it is not picking it up: ‘knee-length’ should be matched with the ‘knee length’ tag under ‘Leg_Length’
Another issue with the dash. Consider the sentence: “This is a mid-weight dress”. Since one of the first steps is to remove the dash, we end up with “this is a mid weight dress”. Then the tagger identifies “mid” as a leg length. However it should not be tagged as such, since it actually is “mid-weight” and not “mid”.

Comments (4)