Table 1.

Results of Amazon_6 in two classes for 4000 texts by k-nearest neighbor (KNN).

Model	Accuracy and F1 of Different Dimensions
	200		300		500		800
	Acc	F1	Acc	F1	Acc	F1	Acc	F1
FW	0.924	0.9239	0.923	0.9229	0.9125	0.9124	0.9003	0.9001
P2	0.9078	0.9075	0.9075	0.9073	0.9078	0.9075	0.9083	0.908
PW	0.9555	0.9554	0.9543	0.9542	0.9548	0.9547	0.955	0.9549
FP2	0.9335	0.9334	0.9273	0.9272	0.9208	0.9207	0.907	0.9069
FPW	0.9593	0.9592	0.9555	0.9549	0.9585	0.9584	0.958	0.9579
FPC	0.9565	0.9564	0.957	0.957	0.9557	0.9557	0.9575	0.9574
Accuracy and F1 of Baselines
Model	Acc				F1
TF–IDF	0.7798				0.7683
LDA	0.954				0.9539
Word2Vec	0.9538				0.9537