摘要：This study uses collaborative filtering approach and proposes an Ant Recommender System (ARS) based on collaborative behavior of ants for generating Top-N recommendations. Present proposed system ARS works in two phases. In the first phase, opinions from users collected in the form of user-item rating matrix are clustered offline using ant based clustering algorithm into predetermined number of clusters and stored in the database for future recommendations. In the second phase, the recommendations are generated online for the active user. The pheromone updating strategy of ants is combined with similarity measure for choosing the clusters with good quality ratings. This helps in improving the quality of recommendations for the active user. The performance of ARS is evaluated using Jester dataset available on the website of University of California, Berkeley and compared with traditional collaborative filtering based recommender system.
