Evaluation - task1-intertass-CR
1233 tweets in qrel, 1233 tweets in run.

Accuracy: 0.560

Confusion matrix
Real	Pred.	#
N	N	370
N	NONE	61
N	P	51
N	NEU	9
NEU	N	79
NEU	P	50
NEU	NONE	34
NEU	NEU	1
NONE	N	78
NONE	NONE	101
NONE	P	40
NONE	NEU	5
P	P	218
P	N	70
P	NONE	59
P	NEU	7

Category evaluation
Cat	P	R	F1
N	0.620	0.754	0.680
NEU	0.045	0.006	0.011
NONE	0.396	0.451	0.422
P	0.607	0.616	0.612

Macro-Precision: 0.417
Macro-Recall: 0.457
Macro-F1: 0.436