Effective Use of Context in Noisy Entity Linking

Mueller, David S.; Durrett, Greg

doi:10.18653/v1/d18-1126

Cited by 14 publications

(21 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This model shows comparable performance to GRU-ATTN, achieving 76.0 accuracy on the original test set of the Wik-ilinksNED data, comparable to the performance of 75.8 reported inMueller and Durrett (2018).…”

supporting

confidence: 61%

“…Note that this similarity is computed using distributed representations while traditional cosine similarity is based on word counts (Hoffart et al 2011). GRU-ATTN Our implementation of the attention-based model introduced in Mueller and Durrett (2018). This model achieves state-of-the-art performance on the WikilinksNED dataset in the standard supervised setting.…”

Section: Baselinesmentioning

confidence: 99%

“…This model achieves state-of-the-art performance on the WikilinksNED dataset in the standard supervised setting. See Mueller and Durrett (2018) for more details. CBOW+WORD2VEC This simpler baseline model 8 uses the pretrained embeddings and a simple bag-of-words mentioncontext encoder.…”

Section: Baselinesmentioning

confidence: 99%

“…Similar to Eshel et al (2017), we use word2vecf to initialize entity embeddings. We compare the context representations and the entity representation by following Mueller and Durrett (2018) Comparison with other systems Table 1 compares assumptions and resources for different systems. The ET4EL model is a so-called local entity linking model, as it only uses a single mention and context, rather than a global model which does collective inference over the whole document (Ratinov et al 2011, Hoffart et al 2011).…”

Section: Baselinesmentioning

confidence: 99%

See 3 more Smart Citations

Fine-Grained Entity Typing for Domain Independent Entity Linking

Onoe

Durrett

2020

AAAI

Self Cite

View full text Add to dashboard Cite

Neural entity linking models are very powerful, but run the risk of overfitting to the domain they are trained in. For this problem, a “domain” is characterized not just by genre of text but even by factors as specific as the particular distribution of entities, as neural models tend to overfit by memorizing properties of frequent entities in a dataset. We tackle the problem of building robust entity linking models that generalize effectively and do not rely on labeled entity linking data with a specific entity distribution. Rather than predicting entities directly, our approach models fine-grained entity properties, which can help disambiguate between even closely related entities. We derive a large inventory of types (tens of thousands) from Wikipedia categories, and use hyperlinked mentions in Wikipedia to distantly label data and train an entity typing model. At test time, we classify a mention with this typing model and use soft type predictions to link the mention to the most similar candidate entity. We evaluate our entity linking system on the CoNLL-YAGO dataset (Hoffart et al. 2011) and show that our approach outperforms prior domain-independent entity linking systems. We also test our approach in a harder setting derived from the WikilinksNED dataset (Eshel et al. 2017) where all the mention-entity pairs are unseen during test time. Results indicate that our approach generalizes better than a state-of-the-art neural model on the dataset.

show abstract

supporting

confidence: 61%

Section: Baselinesmentioning

confidence: 99%

Section: Baselinesmentioning

confidence: 99%

Section: Baselinesmentioning

confidence: 99%

See 2 more Smart Citations

Fine-Grained Entity Typing for Domain Independent Entity Linking

Onoe

Durrett

2020

AAAI

Self Cite

View full text Add to dashboard Cite

show abstract

“…Learned ensembles can also be effective (Rajani et al, 2017). Concept linking has also been applied to bio-medical literature (Dogan et al, 2014;Zheng et al, 2015;Tsai and Roth, 2016;Zhao et al, 2019) and is most similar to the task of entity linking (Dredze et al, 2010;Durrett and Klein, 2014;Gupta et al, 2017;Mueller and Durrett, 2018). Similar to our approach, Choi et al (2016) learn representations of concepts in UMLS.…”

Section: Concept Linkingmentioning

confidence: 89%

Clinical Concept Linking with Contextualized Neural Representations

Schumacher¹,

Mulyar²,

Dredze³

2020

Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics

View full text Add to dashboard Cite

In traditional approaches to entity linking, linking decisions are based on three sources of information -the similarity of the mention string to an entity's name, the similarity of the context of the document to the entity, and broader information about the knowledge base (KB). In some domains, there is little contextual information present in the KB and thus we rely more heavily on mention string similarity. We consider one example of this, concept linking, which seeks to link mentions of medical concepts to a medical concept ontology. We propose an approach to concept linking that leverages recent work in contextualized neural models, such as ELMo (Peters et al., 2018), which create a token representation that integrates the surrounding context of the mention and concept name. We find a neural ranking approach paired with contextualized embeddings provides gains over a competitive baseline (Leaman et al., 2013). Additionally, we find that a pre-training step using synonyms from the ontology offers a useful initialization for the ranker.

show abstract