Recommendation with quantitative implication rules

Association rules based recommendation is one of approaches to develop recommendation systems. However, such systems just focus on binary dataset, whereas many datasets are in the quantitative form. There are many solutions proposed for this problem such as combining the association rules mining with fuzzy logic, binarizing quantitative data, etc. These proposals have contributed to improving the performance of traditional association rules mining, however, they have to deal with the trade-off between the processing performance and the loss of information. In this paper, we propose a new approach to make recommendations based on implication rules. The experimental results show that our proposed solution can be implemented on quantitative dataset well as well as improve the accuracy and performance of the recommendation systems.