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

Accuracy: 0.474

Confusion matrix
Real	Pred.	#
N	NONE	113
N	P	53
N	N	285
N	NEU	21
NEU	NONE	133
NEU	N	114
NEU	P	96
NEU	NEU	24
NONE	NONE	59
NONE	NEU	5
NONE	P	51
NONE	N	44
P	P	309
P	N	53
P	NONE	57
P	NEU	11

Category evaluation
Cat	P	R	F1
N	0.575	0.604	0.589
NEU	0.393	0.065	0.112
NONE	0.163	0.371	0.226
P	0.607	0.719	0.658

Macro-Precision: 0.435
Macro-Recall: 0.440
Macro-F1: 0.437