Neural Cross-Lingual Entity Linking

By Avirup Sil, Gourab Kundu, Radu Florian, Wael Hamza. 2017

105 citations recorded by the source corpus.

A major challenge in Entity Linking (EL) is making effective use of contextual information to disambiguate mentions to Wikipedia that might refer to different entities in different contexts. The problem exacerbates with cross-lingual EL which involves linking mentions written in non-English documents to entries in the English Wikipedia: to compare textual clues across languages we need to compute similarity between textual fragments across languages. In this paper, we propose a neural EL model that trains fine-grained similarities and dissimilarities between the query and candidate document from multiple perspectives, combined with convolution and tensor networks. Further, we show that this English-trained system can be applied, in zero-shot learning, to other languages by making surprisingly effective use of multi-lingual embeddings. The proposed system has strong empirical evidence yielding state-of-the-art results in English as well as cross-lingual: Spanish and Chinese TAC 2015 datasets.

Subfields: Artificial Intelligence, Natural Language Processing, Information Extraction

How researchers see this paper

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

In Artificial Intelligence

1. Sebastian Riedel
2. Fabio Petroni
3. Luke S. Zettlemoyer
4. Hannaneh Hajishirzi
5. Mohit Bansal
6. Yoav Goldberg
7. Graham Neubig
8. Wen-tau Yih

In Natural Language Processing

1. Sebastian Riedel
2. Fabio Petroni
3. Luke S. Zettlemoyer
4. Mohit Bansal
5. Graham Neubig
6. Hannaneh Hajishirzi
7. Eneko Agirre
8. Percy Liang

In Information Extraction

1. Eneko Agirre
2. Ander Barrena
3. Luke S. Zettlemoyer
4. Aitor Soroa Etxabe
5. Fabio Petroni
6. Sebastian Riedel

Citations in the corpus

• Learning Dense Representations for Entity Retrieval (2019, neutral): “Most recent work on entity resolution has focused on training neural network models for the candidate reranking stage (Francis-Landau et al., 2016; Eshel et al., 2017; Yamada et al., 2017a; Gupta et al., 2017; Sil et al., 2018).”
• Scalable Zero-shot Entity Linking with Dense Entity Retrieval (2019, background): “Prior work often used a fixed number of k candidates where k ranges from 5 to 100 (for instance, Yamada et al. (2016) and Ganea and Hofmann (2017) choose k = 30, (Logeswaran et al.”
• Learning Text Representations for 500K Classification Tasks on Named Entity Disambiguation (2018, methodology): “Contrary to systems trained on in-domain data (Cucerzan, 2012; Chisholm and Hachey, 2015; Globerson et al., 2016; Yamada et al., 2016; Sil et al., 2018), ours is trained on Wikipedia and tested out-of-domain.”
• Collective Entity Disambiguation with Structured Gradient Tree Boosting (2018, result): “It is unclear how candidates were pruned, but the entity candidates generated by this system have high recall and low ambiguity, and they contribute to some of the best results reported for AIDA-test (Yamada et al., 2016; Sil et al., 2018).”
• Entity-aware ELMo: Learning Contextual Entity Representation for Entity Disambiguation (2019, methodology): “We also utilize the following basic features introduced by prior work (Yamada et al., 2016, 2017; Sil et al., 2018; Ganea and Hofmann, 2017; Le and Titov, 2018).”