NIST TAC Knowledge Base Population (KBP2020) Entity Discovery and Linking Track

Overview

The goal of TAC-KBP Entity Discovery and Linking (EDL) is to extract mentions of pre-defined entity types from any language, and link (disambiguate and ground) them to the entities in an English knowledge base (KB). In the past several years we have only focused on five major coarse-grained entity types: person (PER), geo-political entity (GPE), location (LOC), organization (ORG) and facility (FAC). Many real world applications in scenarios such as disaster relief and technical support require us to significantly extend our EDL capabilities to a wider variety of entity types (e.g., technical terms, lawsuits, disease, crisis, vehicles, food, biomedical entities). In TAC-KBP2020 we will extend the number of types from five to thousands defined in YAGO. In addition, we will ask human analysts to provide feedback on system output, which will be used to improve each system.