Entity Disambiguation with Freebase

Abstract

Abstract-Entity disambiguation with a knowledge base becomes increasingly popular in the NLP community. In this paper, we employ Freebase as the knowledge base, which contains significantly more entities than Wikipedia and others. While huge in size, Freebase lacks context for most entities, such as the descriptive text and hyperlinks in Wikipedia, which are useful for disambiguation. Instead, we leverage two features of Freebase, namely the naturally disambiguated mention phrases (aka aliases) and the rich taxonomy, to perform disambiguation in an iterative manner. Specifically, we explore both generative and discriminative models for each iteration. Experiments on 2, 430, 707 English sentences and 33, 743 Freebase entities show the effectiveness of the two features, where 90% accuracy can be reached without any labeled data. We also show that discriminative models with proposed split training strategy is robust against overfitting problem, and constantly outperforms the generative ones.

Analysis of named entity recognition and linking for tweets

“…Approaches have used different entity databases as a disambiguation target (e.g. Wikipedia pages [26,27,28]) and Linked Open Data resources (e.g. DBpedia [29], YAGO [30], Freebase [28]).…”

“…Wikipedia pages [26,27,28]) and Linked Open Data resources (e.g. DBpedia [29], YAGO [30], Freebase [28]). Linking entity mentions to such resources is core to semantic web population and identifying differences among linking tools is crucial to achieving harmonious output [31].…”

Discovering Entity Knowledge Bases on the Web

“…Entity linking and wikification have typically relied on Wikipedia (Cucerzan, 2007;Milne and Witten, 2008) or a subset (McNamee et al, 2009), or a larger structured resource such as Freebase (Zheng et al, 2012). Entries in the KB provide a point against which mentions that refer to that entity are clustered.…”

Understandable Big Data: A survey

“…Let us note that if [70] tackle the problem of finding which mentions of entities in a text are equivalent, some authors address a similar problem called Entity Linking and which can be helpful in our disambiguation task. It aims to identify an entry in a given Knowledge Base (KB) to which an entity mention in a document refers to [74,75].…”