rti.htel,2(2):40-55,2009 The success of the collaborative filtering technology depends on the process of locating people with similar profiles (neighbors). We believe that an approach for generating recommendations based a promising research area. In the real ant systems, ants tend to lay a substance called pheromone hile walking from their nests to food source and vice versa. Ants are attracted by pheromones coming from fellow type ants and repulsed by pheromone of non-fellow type ants. Paths that are marked by stronger amount of pheromone are chosen with higher probability than those that have weaker amount of pheromone deposit. This collaborative behavior between identical type ants has an analogy with the collaborative world as people mostly collect opinions from their like-minded friends, neighbors etc One of the main limitations of CF based recommender systems is that it provides recommendations based solely on the opinion of the users whose preferences best matches the taste of active user. A case may arise when an item has not been rated well in the cluster whose similarity matches best with active user but is rated well in other clusters which have similarity closer to his taste. In such a scenario, combinning pheromone information with similarity measure for choosing clusters provides active user with good set of altemative recommendations. In addition to his taste, clusters that are marked by stronger amount of pheromone have the higher probability of being chosen than those that have weaker amount of pheromone deposit. Also, one of the fundamental challenges for recommender systems is to improve the quality of the recommendations. In such a scenario, pheromone updating step guides the search to a better recommendation. The positive feedback in the form of pheromone deposition results in achieving an emergent, unified behavior for the recommender system as a whole and produces a robust system capable of finding improved quality recommendations. In this study, we have tried to improve the quality of recommendation based on the pheromone density and pheromone updating strategy PROPOSED ANT RECOMMENDER SYSTEM Ant Recommender System(ARS)works in two phases. Phase I is the preprocessing phase, which is offline. In this phase, the background data in the form of user-item rating matrix is collected and clustered using an ant based clustering algorithm into predetermined number of clusters. Once the clusters are obtained, the cluster data along with their centroids are stored in the database for future recommendations. Phase II is online in which the recommendation process takes place for the active user. Here, the pheromone deposition/evaporation technique known from ant algorithms is combined with similarity measure for choosing the best clusters for making the recommendations. Rating quality of each item unrated by active user is computed in the chosen clusters. To generate recommendations, clusters are further selected from the chosen clusters based on rating quality of item Recommendations are then made by computing the weighted average of the ratings of items in the elected clusters. Pheromone is updated based on recommendations made to active user and past recommendations. This helps in improving the quality of recommendations for future users. Figure 1 shows the steps where ant colony metaphor has been applied to the traditional collaborative filtering based recommendation process The working of ARS is described below in detail with the help of Jester data set example Phase 1: Preprocessing Phase Step 1: Normalization of Background Data i.e., Rating Matrix User-item ratings taken from jester dataset rated in the scale of-10 to +10 is normalized in the scale 0 to l, where 0 indicates that the item is not rated by the corresponding user. To ease the discussion, running example shown in Table I is used, where U -U,o are users and J, -Jo are items Jokes)rated/ unrated by the user