After annotating the entire dataset in each language, there was an additional annotator for each language who reviewed the entire set. Annotators were volunteers, and theyare acknowledged at theendofthiswork.

For transductivelink prediction, we further propose astochastic embedding layer to model uncertainty in the link prediction between unseen entities.

Weextend our analysis to the more general class of vector symbolic architectures (VSA), which compute withhigh-dimensional vectors(hypervectors) thatarenotnecessarily binary.