I hadn't seen RecDB, thanks for the link! At a glance, other than the licensing, the main point of comparison is that MLDB is more powerful/general-purpose than RecDB, as it provides primitives for not only for creating recommenders but to do other machine-learning tasks like classification, clustering, dimensionality reduction, visualization etc. MLDB also seems to have a bit more documentation :)
I hadn't seen RecDB, thanks for the link! At a glance, other than the licensing, the main point of comparison is that MLDB is more powerful/general-purpose than RecDB, as it provides primitives for not only for creating recommenders but to do other machine-learning tasks like classification, clustering, dimensionality reduction, visualization etc. MLDB also seems to have a bit more documentation :)