WikiGUM: Exhaustive Entity Linking for Wikification in 12 Genres

TL;DR

WikiGUM is a comprehensive dataset for entity linking across 12 diverse genres, including nested and pronominal mentions, enabling improved research and evaluation in Wikification tasks.

Contribution

The paper introduces WikiGUM, a fully annotated, genre-diverse dataset covering all mention types, addressing gaps in existing resources for entity linking research.

Findings

Pretrained SOTA systems perform poorly on the new dataset.

The dataset includes nested, non-named, and pronominal mentions.

Provides a resource for further research on entities in varied contexts.

Abstract

Previous work on Entity Linking has focused on resources targeting non-nested proper named entity mentions, often in data from Wikipedia, i.e. Wikification. In this paper, we present and evaluate WikiGUM, a fully wikified dataset, covering all mentions of named entities, including their non-named and pronominal mentions, as well as mentions nested within other mentions. The dataset covers a broad range of 12 written and spoken genres, most of which have not been included in Entity Linking efforts to date, leading to poor performance by a pretrained SOTA system in our evaluation. The availability of a variety of other annotations for the same data also enables further research on entities in context.