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

Accuracy: 0.324

Confusion matrix
Real	Pred.	#
N	NONE	246
N	P	46
N	N	161
N	NEU	19
NEU	N	78
NEU	P	50
NEU	NONE	224
NEU	NEU	15
NONE	NONE	100
NONE	N	28
NONE	P	23
NONE	NEU	8
P	N	42
P	P	186
P	NONE	189
P	NEU	13

Category evaluation
Cat	P	R	F1
N	0.521	0.341	0.412
NEU	0.273	0.041	0.071
NONE	0.132	0.629	0.218
P	0.610	0.433	0.506

Macro-Precision: 0.384
Macro-Recall: 0.361
Macro-F1: 0.372