Story Generation Using Knowledge Graph under Psychological States

Table 3

Automatic evaluations of the proposed model and the baseline models. The context-merge and context-independent represent different methods of encoder mentioned in [10]. The former is to encode the context and sentence together, while the latter encodes them separately and then concatenates them.

Model	BLEU				ROUGE			METEOR	AGPS
	B-1	B-2	B-3	B-4	R-1	R-2	R-l
Seq2Seq	0.202	0.049	0.016	0.007	0.122	0.014	0.117	0.062	0.654
Transformer	0.207	0.053	0.020	0.008	0.139	0.012	0.128	0.069	0.747
Inc-S2S	0.224	0.053	0.017	0.006	0.151	0.013	0.141	0.067	0.825
SoCP+context-merger	0.216	0.056	0.021	0.010	0.144	0.016	0.136	0.067	0.886
SoCP+context-independent	0.232	0.062	0.025	0.011	0.161	0.018	0.151	0.072	0.879
(Ours)K-GuCCI	0.242	0.070	0.029	0.014	0.169	0.022	0.160	0.077	0.914