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

Accuracy: 0.442

Confusion matrix
Real	Pred.	#
N	P	42
N	NEU	87
N	NONE	117
N	N	226
NEU	NEU	82
NEU	N	83
NEU	NONE	141
NEU	P	61
NONE	P	39
NONE	NONE	62
NONE	N	28
NONE	NEU	30
P	NEU	61
P	NONE	63
P	N	45
P	P	261

Category evaluation
Cat	P	R	F1
N	0.592	0.479	0.529
NEU	0.315	0.223	0.262
NONE	0.162	0.390	0.229
P	0.648	0.607	0.627

Macro-Precision: 0.429
Macro-Recall: 0.425
Macro-F1: 0.427