MIT Press

Skip to Main Content

Table 9

Experimental results for training on the Reddit data set and testing on IAC_v2 using the best LSTM models (sentence-level attention).

Experiment	S			NS
Experiment	P	R	F1	P	R	F1
LSTM^ct_{a_s}	66.51	61.11	63.69	64.03	69.23	66.53
LSTM^ct_{a_s} + LSTM^pt_{a_s}	63.96	60.68	62.28	62.60	65.81	64.17

Experiment	S			NS
Experiment	P	R	F1	P	R	F1
LSTM^ct_{a_s}	66.51	61.11	63.69	64.03	69.23	66.53
LSTM^ct_{a_s} + LSTM^pt_{a_s}	63.96	60.68	62.28	62.60	65.81	64.17