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

Accuracy: 0.256

Confusion matrix
Real	Pred.	#
N	NEU	10
N	NONE	345
N	P	19
N	N	98
NEU	NONE	292
NEU	P	30
NEU	N	32
NEU	NEU	13
NONE	NONE	115
NONE	P	18
NONE	N	22
NONE	NEU	4
P	N	38
P	NONE	246
P	P	140
P	NEU	6

Category evaluation
Cat	P	R	F1
N	0.516	0.208	0.296
NEU	0.394	0.035	0.065
NONE	0.115	0.723	0.199
P	0.676	0.326	0.440

Macro-Precision: 0.425
Macro-Recall: 0.323
Macro-F1: 0.367