The similarity exploited in Table 1.

[needylove]

Dear authors,

Could I know what similarity you exploited in Table 1? The two rank correlation coefficients improved very limited (ranksim even decrease a lot) on other datasets using cos similarity or Euclidean distance as similarity.

Thanks.