I see that for each row there's an answer that says if there's a relation or not and an approval score that tells you how much you should trust that value.
The problem is that many A B pairs have multiple rows with different answers and an approval score of 100%, like these highlighted rows for example:
These two rows contradict each other but at the same time they both are deemed as trusted sources.
Am I reading the dataset wrong? Or how are you supposed to deal with these scenarios?
I see that for each row there's an answer that says if there's a relation or not and an approval score that tells you how much you should trust that value.
The problem is that many A B pairs have multiple rows with different answers and an approval score of 100%, like these highlighted rows for example:
These two rows contradict each other but at the same time they both are deemed as trusted sources.
Am I reading the dataset wrong? Or how are you supposed to deal with these scenarios?