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

Accuracy: 0.389

Confusion matrix
Real	Pred.	#
N	NONE	174
N	P	28
N	N	237
N	NEU	33
NEU	NONE	198
NEU	NEU	26
NEU	N	98
NEU	P	45
NONE	NONE	85
NONE	N	35
NONE	P	24
NONE	NEU	15
P	N	69
P	NONE	130
P	P	207
P	NEU	24

Category evaluation
Cat	P	R	F1
N	0.540	0.502	0.520
NEU	0.265	0.071	0.112
NONE	0.145	0.535	0.228
P	0.681	0.481	0.564

Macro-Precision: 0.408
Macro-Recall: 0.397
Macro-F1: 0.402