Towards solution-oriented knowledge recommendation

Abstract

This research identifies three practical problems in the existing scholarly recommendation systems, which leads to three research issues related to solution-oriented scientific knowledge recommendation that have not been addressed in the literature, including the lack of solution-oriented article recommendation, the lack of solution-oriented knowledge repositories, and the lack of weighted bibliometric network for ranking academic articles. Individual research could be conducted to address each issue, however this study proposes a knowledge fusion framework to collectively integrate multiple models which are specifically designed to solve each one of the issues respectively. The framework will be able to generate valuable knowledge repositories containing the solutions proposed in the mass scholarly articles to answer academic questions, and meanwhile it can rank these solutions based on weighted bibliometric networks.