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

Accuracy: 0.324

Confusion matrix
Real	Pred.	#
N	NONE	252
N	P	62
N	N	122
N	NEU	36
NEU	NONE	225
NEU	N	45
NEU	NEU	26
NEU	P	71
NONE	P	30
NONE	NEU	13
NONE	NONE	101
NONE	N	15
P	NONE	156
P	P	214
P	N	37
P	NEU	23

Category evaluation
Cat	P	R	F1
N	0.557	0.258	0.353
NEU	0.265	0.071	0.112
NONE	0.138	0.635	0.226
P	0.568	0.498	0.530

Macro-Precision: 0.382
Macro-Recall: 0.366
Macro-F1: 0.374