implementing relational inner product association (RIPA) #7
Description
This paper claims that one can hack WEAT by cherry-picking words in A and B. The RIPA can protect against such hacking.
The method RIPA does not appear to be difficult to implement. But the fact that the paper doesn't publish any data bothers me.
Words in A and B must be in pair, e.g.
A <- c("man", "men", "king")
B <- c("woman", "women", "queen")