Open knowledge graph completion with negative-aware representation learning and multi-source reliability inference

Huang Peng, Weixin Zeng, Jiuyang Tang, Mao Wang, Hongbin Huang, Xiang Zhao

Abstract: Highlights•We propose a novel solution towards open knowledge graph completion.•We use negative sampling to facilitate recognition of false claims.•We consider both prior knowledge and quality of sources during inference.•Applications on multi-source data fusion for sustainable cities are investigated.