rti.htel,2(2):40-55,2009 Table 9: Recommendations generated for active user Clusters chosen 6) Where Quality of item in the cluster selected No of clusters selected avg rating Average rating of the item in the selected cluster The computed ratings are shown in Table 9. If the number of clusters selected is more than one, then rating is computed as the weighted average, otherwise rating is computed as the average rating of the item in the cluster selected. For joke 2, avg rating is computed from cluster 2, for jokes 4 and 9, avg_ rating is computed from cluster 3 and for joke 6, weighted avg rating is computed from clusters 2 and 3 Step 4: Pheromone Updating The aim of the pheromone updating strategy is to increase the pheromone values associated with ood solutions and to decrease those that are associated with bad ones. Usually, this is achieved by decreasing all the pheromone values through pheromone evaporation and by increasing the pheromone levels associated with a chosen set of good solutions. This is analogous to the phenomenon of pheromone evaporation and deposition in real ant colonies. As shown in Eq. 7, the pheromone associated to each chuster is decreased by a small value and pheromone associated to the cluster from hich recommendation is generated is increased proportional to the rating quality(aQ) of the item The amount of pheromone associated with each cluster i updated is given Phc(t)=0-p)×Pher(t-1)+△Q×Pher(t-1) Where Qe if cluster selected>1 △Q otherwise Where Pheromone evaporation rate, large value of p indicates a fast evaporation and vice versa Qcc- Rating quality of item in the selected cluster No. of clusters selected As shown in Table 9, recommendation for joke 3 is generated from cluster 2, joke 4 and 9 are generated from cluster 3 and recommendation for joke 6 is generated from clusters 2 and 3. Therefore