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

Accuracy: 0.556

Confusion matrix
Real	Pred.	#
N	N	338
N	P	66
N	NEU	35
N	NONE	52
NEU	N	61
NEU	NEU	31
NEU	NONE	29
NEU	P	43
NONE	N	73
NONE	NONE	95
NONE	NEU	14
NONE	P	42
P	P	221
P	N	67
P	NEU	24
P	NONE	42

Category evaluation
Cat	P	R	F1
N	0.627	0.688	0.656
NEU	0.298	0.189	0.231
NONE	0.436	0.424	0.430
P	0.594	0.624	0.609

Macro-Precision: 0.489
Macro-Recall: 0.481
Macro-F1: 0.485