The table below presents the human evaluation for food recommendation systems. It consists of presenting a set of evaluators with questions from a randomized test, along with the user personas spanning ingredient preferences (likes and dislikes) and applicable nutrition guidelines. This table shows that the personalized baseline models (prefixed by P-) outperform the non-personalized BAMNet model. On the other hand, the pFoodREQ model proposed by Yu et al. utilizing personal information such as dietary preferences and health guidelines to provide more accurate and personalized food information outperforms the baselines models.
Created by
Comparison update
ORKG Comparisons have changed. We have added new features and improved the user interface. Comparisons might look slightly different, but the comparison data itself remains unchanged.