In this paper we evaluate the performance of different collaborative algorithms in cold-start situations, where the initial lack of ratings may affect the quality of the algorithms.
Instead of trying to design an algorithm that provides, on average, a reasonable good accuracy in both the early and steady stages of a recommender system, in this paper we suggest adopting different algorithms in the different stages of the recommender system life-cycle. The evaluation has been performed on the datasets collected by two digital-television providers in Europe. Both the datasets have been implicitly collected by analyzing the pay-per-view movies purchased by the users over a period of several months.
The first result of the paper outlines that item-based algorithms perform better with respect to SVD-based ones in the early stage of the cold-start problem. The second result shows that the accuracy of SVD-based algorithms, when using few latent factors, decreases with the time-evolution of the dataset. On the contrary, SVD-based algorithms, when used with a large-enough number of latent features, increase theirs accuracy with time and may outperform the item-based algorithms if the dataset does not present a long-tail behavior.
PDF, 14 pages, 0.6 MB