McFarland Punch list

McFarland Punch list NEXT UP computational literacy .com

1. Reframe the final paper argument

2. Get final analyses into the final paper.

3. Add descriptive statistics comparing the FATML community with the mainstream ML community. 

4. Arrange all the analysis into the final argument. 

5. Loose ends to check and tie off:  Look at whether we can make different inferences based on US/Asian. Asian, White, URM, unknown Checking

6. Can we improve the international stuff?

7. Teams that have ANY representation?

8. What if we use percentage instead of cutoff for classification. 

9. Do different thirds load differently?

10. Following Vinod and Jurgen’s work, position in paper predicts whether it’s rising or dying. 

Do we find a topic for fairness? 

The fairness topic

I want to be able to say whether there’s a historical fairness topic.  I’m currently trying to get a dataframe of metadata that aligns with the GuidedLDA corpus. 

WHAT’s NEXT?

Once I have the qualitative stuff for the GuidedLDA:

— write functions to generate my regular graphs — descriptive bit.