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

Accuracy: 0.451

Confusion matrix
Real	Pred.	#
N	N	270
N	NONE	140
N	P	30
N	NEU	32
NEU	NONE	216
NEU	P	57
NEU	N	62
NEU	NEU	32
NONE	NONE	92
NONE	N	29
NONE	P	31
NONE	NEU	7
P	P	250
P	NONE	125
P	NEU	17
P	N	38

Category evaluation
Cat	P	R	F1
N	0.677	0.572	0.620
NEU	0.364	0.087	0.141
NONE	0.161	0.579	0.251
P	0.679	0.581	0.627

Macro-Precision: 0.470
Macro-Recall: 0.455
Macro-F1: 0.462