Entity Linking at Web Scale

By Thomas Lin, Mausam, Oren Etzioni. 2012

137 citations recorded by the source corpus.

Subfields: Information Retrieval & Search, Natural Language Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Information Retrieval & Search

In Natural Language Processing

Citations in the corpus

Semantic Parsing on Freebase from Question-Answer Pairs (2013, methodology): “Specifically, when we combine logical forms z1 and z2 via a join or bridging, we include a feature on the POS tag of (the first word spanned by) z1 conjoined with the POS tag corresponding to z2.”
Large-scale Simple Question Answering with Memory Networks (2015, methodology): “To link the subject and the object of a Reverb fact to Freebase entities, we use precomputed entity links (Lin et al., 2012).”
Semantic Parsing via Paraphrasing (2014, methodology): “For Free-base entities, we extract a popularity feature based on the entity frequency in an entity linked subset of Reverb (Lin et al., 2012).”
Question Answering over Freebase with Multi-Column Convolutional Neural Networks (2015, methodology): “Specifically, we kept the triples where one of the entities appeared in the training/development set of WEBQUESTIONS or CLUEWEB extractions provided in (Lin et al., 2012), and removed the entities appearing less than five times.”
Open question answering over curated and extracted knowledge bases (2014, methodology): “The algorithm maintains a beam of partial derivations, each scored by the function score(d|f ,w).”
Constrained Semi-supervised Learning in the Presence of Unanticipated Classes (2016, background): “In fact, we face a ‘chicken and egg’ problem, as most EL methods assume that KB entities have glosses, whereas we do not [60, 80].”
Knowledge-Based Question Answering as Machine Translation (2014, methodology): “Berant et al. (2013) have used a lexicon extracted from a subset of ReVerb triples (Lin et al., 2012 P recision is deﬁned as the number of correctly answered questions divided by the number of questions with non-empty answers generated by our KB-QA system.”