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

Accuracy: 0.173

Confusion matrix
Real	Pred.	#
N	NONE	393
N	P	72
N	NEU	5
N	N	2
NEU	NONE	290
NEU	P	72
NEU	N	3
NEU	NEU	2
NONE	NONE	118
NONE	P	41
P	NONE	303
P	P	125
P	N	2

Category evaluation
Cat	P	R	F1
N	0.286	0.004	0.008
NEU	0.286	0.005	0.011
NONE	0.107	0.742	0.187
P	0.403	0.291	0.338

Macro-Precision: 0.270
Macro-Recall: 0.261
Macro-F1: 0.265