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

Accuracy: 0.352

Confusion matrix
Real	Pred.	#
N	P	66
N	NONE	209
N	N	142
N	NEU	55
NEU	NONE	192
NEU	N	61
NEU	NEU	37
NEU	P	77
NONE	NONE	99
NONE	NEU	9
NONE	P	30
NONE	N	21
P	N	29
P	P	225
P	NONE	150
P	NEU	26

Category evaluation
Cat	P	R	F1
N	0.561	0.301	0.392
NEU	0.291	0.101	0.150
NONE	0.152	0.623	0.245
P	0.565	0.523	0.543

Macro-Precision: 0.393
Macro-Recall: 0.387
Macro-F1: 0.390