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

Accuracy: 0.501

Confusion matrix
Real	Pred.	#
N	N	373
N	NEU	14
N	P	36
N	NONE	49
NEU	N	178
NEU	P	77
NEU	NONE	103
NEU	NEU	9
NONE	NONE	52
NONE	N	65
NONE	P	39
NONE	NEU	3
P	P	281
P	N	103
P	NONE	35
P	NEU	11

Category evaluation
Cat	P	R	F1
N	0.519	0.790	0.626
NEU	0.243	0.025	0.045
NONE	0.218	0.327	0.261
P	0.649	0.653	0.651

Macro-Precision: 0.407
Macro-Recall: 0.449
Macro-F1: 0.427