Closed ardate closed 5 years ago
Thanks! Sorry for the delay. This seems to be better called "containment min", because it would be the minimum of two containment scores |x \cap y| / |x|
and |x \cap y| / |y|
. "containment max" would be dividing by min(|x|, |y|)
.
Hi Eric,
Thanks for your message! Indeed you are right.
Best, Arda
On 21 Mar 2019, at 20:33, Eric Zhu notifications@github.com wrote:
Thanks! Sorry for the delay. This seems to be better called "containment min", because it would be the minimum of two containment scores |x \cap y| / |x| and |x \cap y| / |y|. "containment max" would be dividing by min(|x|, |y|).
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/ekzhu/SetSimilaritySearch/pull/2#issuecomment-475372254, or mute the thread https://github.com/notifications/unsubscribe-auth/AG50czk0-T4IJ8hM9PmtBr5ConP6Wcgeks5vY95ugaJpZM4bFM3r.
Based on our communication via e-mail, I am adding a new similarity function "containment_max". This similarity function is similar to "containment", with the difference that the similarity value between two sets is normalised by the max. length of the two sets (instead of normalising it by the length of set1 as in "containment" function).