Abstract
In order to answer natural language questions over knowledge graphs, mostprocessing pipelines involve entity and relation linking. Traditionally, entitylinking and relation linking has been performed either as dependent sequentialtasks or independent parallel tasks. In this paper, we propose a framework,calledEARL, which performs entity linking and relation linking as a jointsingle task. EARL is modeled on an optimised variation of GeneralisedTravellingSalesperson Problem. The system determines the best semantic connection betweenall keywords of the question by referring to the knowledge graph. This isachieved by exploiting the connection density between entity candidates andrelation candidates. We have empirically evaluated the framework on a datasetwith 3000 complex questions. Our system surpasses state-of-the-art scores forentity linking task by reporting an accuracy of 0.67against 0.40 from the nextbest entity linker