Academic journal recommendation for human neuroimaging studies via brain activation-based filtering

The development of noninvasive neuroimaging techniques such as functional magnetic resonance imaging (fMRI) was followed by a sheer volume of human neuroimaging studies to understand mental processes, mechanisms, and diseases. Due to the information overload, it is increasingly challenging for neuroscientists to review existing academic journals and find the most suitable journal to publish their studies. Therefore, this paper proposes an academic journal recommendation model for human neuroimaging studies called brain activation-based filtering (BAF), which predicts relevant journals even before papers are written. In particular, BAF utilizes a collective matrix factorization (CMF) model that predicts relevant journals based on brain regions activated in specific neuroimaging studies. The proposed model predicts journals of published human neuroimaging studies with a reliable AUC score of 0.856 by implementing five-fold cross-validation. The implementation code is publicly available at https://github.com/JunsolKim/brain-activation-based-filtering.