Title: "An overview on Entity discovery and linking systems"

Abstract:
"Entity linking is the task of disambiguating nominal mentions by connecting them to their corresponding concept in a knowledge base. Concept clustering is the task of grouping the mentions that denote the same concept. In the case of ambiguous proper names, we apply reference resolution and clustering of those with an identical reference. We will provide an overview of the main components of an entity discovery and linking system and revisit the design challenges involved in it. Also we will look at the relation between the tasks and how they can complement each other. Finally we will go through the different techniques that have been studied in recent work to tackle these problems, and propose a deep learning approach with special emphasis on multilingual cross-document entity linking."