Algorithm 4: Two-step QABC clustering

Input: D dataset contains best feature values $f_{best,i}$ for each ${URL}_i$ Fitness Function “Sensitivity” Output: Best solution of final cluster center ($C_{best,j}$) j = 1, 2 Begin Initialization phase. -Important QABC settings are determined by the user, including particle population size, optimization variable bounds, modification rate (MR), tuning vector ($\overrightarrow{\mu }$), and stopping criterion (t_max). -Use a uniform probability distribution function to seed an initial population (array) of food sources with placements across the n-dimensional space. For i = 1:SN/* SN is the total number of food sources (number of clusters) */ Initialize the food source within the boundary of given dataset in random order; Apply the K-means algorithm Send the employed bees to the food sources; /* Computed centers */ End For Iteration = 0; Do While (the termination condition is not met) For (each employed bee)/* Employed bee’s phase */ For i = 1:SN -Update the positions of the employed bees using $v_{ij}=x_{ij}+{\varnothing }_{ij}\left(x_{ij}-x_{kj}\right),{\varnothing }_{ij}\in [-\mathrm{1,1}]$, then evaluate their fitness using Equation (3) -A greedy selection is made between the old and new food source and keep the best one; End For For i = 1:SN Compute the probability value associated with each food source.  End For For i = 1:SN /* Start the onlooker bees phase*/  If (rand ( ) < Pri) /* Pri the probability associated with ith food source */ -Generate a new position for each onlooker using Equation (15) then evaluate the fitness -A greedy selection is made between the old and new food source and keep the best one; Else i = i + 1; End If End For If (any employed bee becomes scout bee)/* Scout bee’s phase */ Send the scout bee to a randomly produced to food source; End if Memorize the best solution achieved Iteration = iteration + 1 End While Obtain final cluster Center End