About the Coding Exercise 1 Results

Master / Vertiefung Nailing your Thesis
1

Disclaimer: Dieser Thread wurde aus dem alten Forum importiert. Daher werden eventuell nicht alle Formatierungen richtig angezeigt. Der ursprüngliche Thread beginnt im zweiten Post dieses Threads.

2

About the Coding Exercise 1 Results
Can you please explain what do the F-measure, Recall, and Precision mean? How should we interpret the results if we did good or not? I think only precise seems clear, but I wasn’t able to interpret others, and I am sorry if it is explained in the class.

Thank you in advance.

3

i think this applies: https://en.wikipedia.org/wiki/F1_score

4

Indeed. Thanks, Ferron.

Also to give a bit of interpretative context w.r.t. the exercise: we consider agreement above 20% agreement good, and everything above 40% excellent.

5

Hi Andreas,

short question regarding the correction: is there a way to look up mistakes that one made while tagging?

Thanks in advance,
Florian

6

There is not, since we’ve been using the same data set for multiple iterations of NYT, and revealing the „optimal“ coding would likely have an impact on future iterations of NYT.
Sorry.