Open elizlee opened 2 years ago
Currently we are only querying an event's PB label for event qnodes. We did this because we believed that this would reduce odd results, but I wonder how determining the best result between the source text and the PB label will impact the output.
Currently we are only querying an event's PB label for event qnodes. We did this because we believed that this would reduce odd results, but I wonder how determining the best result between the source text and the PB label will impact the output.