HIM-PRS: A Patent Recommendation System Based on Hierarchical Index-Based MapReduce Framework

摘要

Intellectual Property (IP) data, such as patent documents, grows inconceivably in recent years. Therefore, discovering valuable information from those huge number of data becomes a challenge. This paper introduces a novel patent recommendation system called HIM-PRS which is built on top of hierarchical index based big data processing platform. HIM-PRS integrates with linked data to provide an efficient patent recommendation service. Our result shows that HIM-PRS is able to find more semantically similar patents than other systems. Additionally, HIM-PRS launches query jobs at least 2 times faster than original Hadoop MapReduce framework.