Stratification-Based Outlier Detection over the Deep Web

Algorithm 2

.

(1) pilot sample ← initial random sample

(2) 

(3) 

(4

(5) compute , the probability of containing outlier for each

(6)  for all    do

(7)  

(8)   data records sampled from

(9)  

(10) end for

(11)

(12)

(13) for all    do

(14)  if  (DR) >   then

(15)   

(16)  else if  (DR) ≥   then

(17)   US ← US + DR

(18)  end if

(19) end for

(20) for all    do

(21)  

(22)   data records sampled from

(23)  

(24) end for

(25) for all  DR

(26)  if  (DR) >   then

(27)   

(28)  end if

(29) end for