Evaluation - task1-intertass-PE
1428 tweets in qrel, 1428 tweets in run.

Accuracy: 0.279

Confusion matrix
Real	Pred.	#
N	NONE	239
N	P	36
N	NEU	69
N	N	128
NEU	NONE	220
NEU	NEU	65
NEU	N	54
NEU	P	28
NONE	P	13
NONE	NONE	107
NONE	N	19
NONE	NEU	20
P	NEU	76
P	NONE	222
P	N	34
P	P	98

Category evaluation
Cat	P	R	F1
N	0.545	0.271	0.362
NEU	0.283	0.177	0.218
NONE	0.136	0.673	0.226
P	0.560	0.228	0.324

Macro-Precision: 0.381
Macro-Recall: 0.337
Macro-F1: 0.358