[1808.07699] End-to-End Neural Entity Linking

Entity Linking (EL) is an essential task for semantic text understanding and information extraction. Popular methods separately address the Mention Detection (MD) and Entity Disambiguation (ED) stages of EL, without leveraging their mutual dependency. We here propose the first neural end-to-end EL system that jointly discovers and links entities in a text document. The main idea is to consider all possible spans as potential mentions and learn contextual similarity scores over their entity candidates that are useful for both MD and ED decisions. Key components are context-aware mention embeddings, entity embeddings and a probabilistic mention - entity map, without demanding other engineered features. Empirically, we show that our end-to-end method significantly outperforms popular systems on the Gerbil platform when enough training data is available. Conversely, if testing datasets follow different annotation conventions compared to the training set (e.g. queries/ tweets vs news docume

1 mentions: @shunsugiyama360
Keywords: end-to-end
Date: 2019/08/27 21:47

Referring Tweets

@shunsugiyama360 t.co/3zgYWin70M 以前この論文と似たことをやるためにOusiaの公開しているentity vectorを使った。特徴量設計をゴリゴリやるモデルよりもエレガントな気はする。

Related Entries

Read more Building Cross-Lingual End-to-End Product Search with Tensorflow · Han Xiao Tech Blog - Deep Learnin...
5 users, 1 mentions 2019/04/08 11:15
Read more Riku Arakawa on Twitter: "現在のend-to-endの手法はニューラルネットすごいのねという感情が強かったから、これぞようやくAIだという感覚になる論文だった。deepmin...
0 users, 0 mentions 2018/07/10 00:23
Read more IBM Research advances in end-to-end speech recognition at INTERSPEECH 2019 | IBM Research Blog
1 users, 4 mentions 2019/10/18 15:09
Read more [1604.04562] A Network-based End-to-End Trainable Task-oriented Dialogue System
0 users, 1 mentions 2019/03/23 11:15