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

Accuracy: 0.144

Confusion matrix
Real	Pred.	#
N	NONE	435
N	N	34
N	P	3
NEU	NONE	346
NEU	N	18
NEU	P	3
NONE	NONE	151
NONE	N	7
NONE	P	1
P	NONE	384
P	N	26
P	P	20

Category evaluation
Cat	P	R	F1
N	0.400	0.072	0.122
NEU	0.000	0.000	0.000
NONE	0.115	0.950	0.205
P	0.741	0.047	0.088

Macro-Precision: 0.314
Macro-Recall: 0.267
Macro-F1: 0.289