Questions about paper

[kangkang59812]

In your paper Section3.1 , you said Adding edge information to already given head or tail information provides minimal gain.. But from Fig.3, p(head | tail, edge) is gained about 5% than p(head | tail) (also the gap between p(tail | head edge) and p(tail | head) is even about 8%). So the gain is still minimal ? Or I read the Fig incorrectly. Thanks!

[fuyun1107]

In your paper Section3.1 , you said Adding edge information to already given head or tail information provides minimal gain.. But from Fig.3, p(head | tail, edge) is gained about 5% than p(head | tail) (also the gap between p(tail | head edge) and p(tail | head) is even about 8%). So the gain is still minimal ? Or I read the Fig incorrectly. Thanks!

have you slove this?xd