Leveraging Trust Relations to Improve Academic Patent Recommendation

Abstract

Academic patent trading is one of the important ways for university technology transfer. Compared to industry patent trading, academic patent trading suffers from a more serious information asymmetric problem. It needs a recommendation service to help companies identify academic patents that they want to pay. However, existing recommendation approaches have limitations in facilitating academic patent trading in online patent platforms because most of them only consider patent-level characteristics. A high trust degree of a company towards academic patents can alleviate the information asymmetry and encourage trading. This study proposes a novel academic patent recommendation approach with a hybrid strategy, combining citation-based relevance, connectivity, and trustworthiness. An offline experiment is conducted to evaluate the performance of the proposed recommendation approach. The results show that the proposed method performs better than the baseline methods in both accuracy and ranking.