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

Accuracy: 0.385

Confusion matrix
Real	Pred.	#
N	NEU	242
N	P	45
N	NONE	67
N	N	118
NEU	NEU	188
NEU	N	44
NEU	NONE	80
NEU	P	55
NONE	NEU	73
NONE	NONE	41
NONE	P	26
NONE	N	19
P	N	44
P	NEU	132
P	NONE	51
P	P	203

Category evaluation
Cat	P	R	F1
N	0.524	0.250	0.339
NEU	0.296	0.512	0.375
NONE	0.172	0.258	0.206
P	0.617	0.472	0.535

Macro-Precision: 0.402
Macro-Recall: 0.373
Macro-F1: 0.387