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

Accuracy: 0.410

Confusion matrix
Real	Pred.	#
N	NEU	191
N	P	43
N	NONE	86
N	N	152
NEU	NEU	158
NEU	N	56
NEU	NONE	106
NEU	P	47
NONE	NONE	57
NONE	NEU	60
NONE	P	27
NONE	N	15
P	N	32
P	NEU	135
P	P	218
P	NONE	45

Category evaluation
Cat	P	R	F1
N	0.596	0.322	0.418
NEU	0.290	0.431	0.347
NONE	0.194	0.358	0.252
P	0.651	0.507	0.570

Macro-Precision: 0.433
Macro-Recall: 0.405
Macro-F1: 0.418