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

Accuracy: 0.471

Confusion matrix
Real	Pred.	#
N	P	43
N	N	318
N	NONE	105
N	NEU	6
NEU	NONE	152
NEU	N	126
NEU	P	77
NEU	NEU	12
NONE	NONE	75
NONE	NEU	4
NONE	P	37
NONE	N	43
P	P	268
P	N	66
P	NONE	83
P	NEU	13

Category evaluation
Cat	P	R	F1
N	0.575	0.674	0.620
NEU	0.343	0.033	0.060
NONE	0.181	0.472	0.261
P	0.631	0.623	0.627

Macro-Precision: 0.432
Macro-Recall: 0.450
Macro-F1: 0.441