Collaborative Semantic Association Discovery from Linked Data
Qing-zhao Zheng, Huajun Chen, Yu Tong, Gang Pan · 2009
The efforts of publishing and interlinking structured data on the Semantic Web will result in a global network of databases, or the Linked Data, which provides huge potential for discovering hidden relationships. We present a multi-agent framework for Semantic Associations Discovery (SAD) from distributed linked data on the Semantic Web. Here, agents collaborate in SAD by publishing inter-dependent hypotheses and evidences, giving rise to an evidentiary network that connects and ranks diverse knowledge elements. We evaluate this framework through simulation, and the results show that the framework is suitable in cross-domain relationship discovery tasks.