Key Features


Data infrastructures make it convenient to work with auxiliary information and enable seamless cross-modality comparisons.


A rich collection of iterators for easy stochastic optimization. Model implementations make use of Cython to achieve C/C++ performance.


Full control over random number generators, open-access to existing algorithms and built-in datasets for reproducible research.


If you use Cornac in a scientific publication, we would appreciate citations to the following paper:

Cornac: A Comparative Framework for Multimodal Recommender Systems, Salah et al., JMLR 21, pp. 1-5, 2020.

Bibtex entry:

  author  = {Aghiles Salah and Quoc-Tuan Truong and Hady W. Lauw},
  title   = {Cornac: A Comparative Framework for Multimodal Recommender Systems},
  journal = {Journal of Machine Learning Research},
  year    = {2020},
  volume  = {21},
  number  = {95},
  pages   = {1-5},
  url     = {}