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

Accuracy: 0.340

Confusion matrix
Real	Pred.	#
N	NONE	282
N	P	24
N	N	166
NEU	NONE	262
NEU	N	72
NEU	P	31
NEU	NEU	2
NONE	NONE	114
NONE	P	21
NONE	N	24
P	N	40
P	NONE	187
P	P	203

Category evaluation
Cat	P	R	F1
N	0.550	0.352	0.429
NEU	1.000	0.005	0.011
NONE	0.135	0.717	0.227
P	0.728	0.472	0.573

Macro-Precision: 0.603
Macro-Recall: 0.387
Macro-F1: 0.471