Evaluation - task1-intertass-PE
3000 tweets in qrel, 1428 tweets in run.

Accuracy: 0.210

Confusion matrix
Real	Pred.	#
-	?	72
N	?	348
N	NONE	128
N	NEU	85
N	P	38
N	N	221
NEU	?	227
NEU	NONE	164
NEU	NEU	68
NEU	P	72
NEU	N	63
NONE	?	599
NONE	NONE	75
NONE	NEU	22
NONE	P	36
NONE	N	26
P	?	326
P	N	43
P	P	265
P	NONE	63
P	NEU	59

Category evaluation
Cat	P	R	F1
-	0.000	0.000	0.000
N	0.626	0.270	0.377
NEU	0.291	0.114	0.164
NONE	0.174	0.099	0.126
P	0.645	0.351	0.454

Macro-Precision: 0.347
Macro-Recall: 0.167
Macro-F1: 0.225