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

Accuracy: 0.358

Confusion matrix
Real	Pred.	#
N	P	52
N	NONE	220
N	N	157
N	NEU	43
NEU	NONE	227
NEU	N	60
NEU	P	55
NEU	NEU	25
NONE	NONE	114
NONE	N	15
NONE	P	26
NONE	NEU	4
P	P	215
P	NONE	166
P	N	25
P	NEU	24

Category evaluation
Cat	P	R	F1
N	0.611	0.333	0.431
NEU	0.260	0.068	0.108
NONE	0.157	0.717	0.257
P	0.618	0.500	0.553

Macro-Precision: 0.411
Macro-Recall: 0.404
Macro-F1: 0.408