Table 1.

Machine learning system evaluation on 500 randomly selected tweets from the unannotated tweets corpus

Category				Precision	Recall	F measure
Overall	Micro-averaging			0.7860	0.7860	0.7860
	Macro-averaging			0.7112	0.7051	0.7081
Per Category	Unrelated			0.9337	0.9482	0.9409
	Related	Positive		0.6596	0.8774	0.7530
		Neutral		0.7586	0.5690	0.6502
		Negative	Safety	0.8039	0.6308	0.7069
			Others	0.4000	0.5000	0.4444