Fig. 7From: A method for evaluating discoverability and navigability of recommendation algorithmsSuccess ratios for the berrypicking scenario. The bars depict the average percentage of found targets (genre-based clusters in the first row and rating-based clusters in the second row). Every simulation consisted of three target clusters, and the success rates are computed as the average number of found targets. Recommendation networks generated by interpolation weights (IW) and matrix factorization (MF) performed bestBack to article page