What evaluation metric is being used for ABSA Task 5 Subtask 1 Slot 1?

[Stefan Falk]

I wonder, what metric is being used for the evaluation of the results? I was hoping that I can do it without the A.jar evaluation application and just calculate the score by myself.

But which metric is being used? Is it just the plain average of all F1-scores for each class or does A.jar calculate one of micro, macro or weighted F1 scores for the final evaluation? 

I could not find anything in the papers or in the provided material.

Thanks in advance!

BR; Stefan

[1]: http://alt.qcri.org/semeval2016/task5/