摘要: Linked Data has emerged as a powerful way of interconnecting structured data on the Web. However, cross-linkage between sources is not extensive one would hope for. In this paper, we formalize task automatically creating "sameAs" links across in globally consistent manner. Our algorithm, presented multi-core well distributed version, achieves link generation by accounting for joint evidence match. Experiments confirm that our system scales beyond 100 million entities and delivers highly accurate results despite vast heterogeneity daunting scale.