• Corpus ID: 232168881

Fast and Effective Biomedical Entity Linking Using a Dual Encoder

  title={Fast and Effective Biomedical Entity Linking Using a Dual Encoder},
  author={Rajarshi Bhowmik and Karl Stratos and Gerard de Melo},
Biomedical entity linking is the task of identifying mentions of biomedical concepts in text documents and mapping them to canonical entities in a target thesaurus. Recent advancements in entity linking using BERT-based models follow a retrieve and rerank paradigm, where the candidate entities are first selected using a retriever model, and then the retrieved candidates are ranked by a reranker model. While this paradigm produces state-of-the-art results, they are slow both at training and test… 

Figures and Tables from this paper

Cross-Domain Data Integration for Named Entity Disambiguation in Biomedical Text

This work proposes a cross-domain data integration method that transfers structural knowledge from a general text knowledge base to the medical domain and achieves state-of-the-art performance on two benchmark medical NED datasets: MedMentions and BC5CDR.

HPI-DHC @ BioASQ DisTEMIST: Spanish Biomedical Entity Linking with Pre-trained Transformers and Cross-lingual Candidate Retrieval

The goal of the task is to extract disease mentions from Spanish clinical case reports and map them to concepts in SNOMED CT and a detailed analysis of system performance highlights the importance of task-specific entity ranking and the benefits of cross-lingual candidate retrieval.

Generative Biomedical Entity Linking via Knowledge Base-Guided Pre-training and Synonyms-Aware Fine-tuning

This work uses a generative approach to model biomedical EL and proposes to inject synonyms knowledge in it, and achieves state-of-the-art results on several biomedical EL tasks without candidate selection.

Seeking Common but Distinguishing Difference, A Joint Aspect-based Sentiment Analysis Model

This work proposes a joint ABSA model, in which a pair encoder especially focuses on candidate aspect-opinion pair classification, and the original encoder keeps attention on sequence labeling, and introduces a dual-encoder design.

Knowledge-Aware Neural Networks for Medical Forum Question Classification

A novel medical knowledge-aware BERT-based model (MedBERT) is developed that explicitly gives more weightage to medical concept-bearing words, and utilize domain-specific side information obtained from a popular medical knowledge base.



LATTE: Latent Type Modeling for Biomedical Entity Linking

LATTE is proposed, a LATent Type Entity Linking model, that improves entity linking by modeling the latent fine-grained type information about mentions and entities and achieves significant performance improvements over several state-of-the-art techniques.

MedType: Improving Medical Entity Linking with Semantic Type Prediction

This paper presents MedType, a fully modular system that prunes out irrelevant candidate concepts based on the predicted semantic type of an entity mention, and incorporates it into five off-the-shelf toolkits for medical entity linking and demonstrates that it consistently improves entity linking performance across several benchmark datasets.

End-to-End Neural Entity Linking

This work proposes the first neural end-to-end EL system that jointly discovers and links entities in a text document and shows that it significantly outperforms popular systems on the Gerbil platform when enough training data is available.

Scalable Zero-shot Entity Linking with Dense Entity Retrieval

This paper introduces a simple and effective two-stage approach for zero-shot linking, based on fine-tuned BERT architectures, and shows that it performs well in the non-zero-shot setting, obtaining the state-of-the-art result on TACKBP-2010.

Learning Dense Representations for Entity Retrieval

We show that it is feasible to perform entity linking by training a dual encoder (two-tower) model that encodes mentions and entities in the same dense vector space, where candidate entities are

BERT-based Ranking for Biomedical Entity Normalization

  • Zongcheng JiQiang WeiHua Xu
  • Computer Science
    AMIA Joint Summits on Translational Science proceedings. AMIA Joint Summits on Translational Science
  • 2020
The experimental results show that the best fine-tuned models consistently outperformed previous methods and advanced the state-of-the-art for biomedical entity normalization, with up to 1.17% increase in accuracy.

Zero-Shot Entity Linking by Reading Entity Descriptions

It is shown that strong reading comprehension models pre-trained on large unlabeled data can be used to generalize to unseen entities and proposed domain-adaptive pre-training (DAP) is proposed to address the domain shift problem associated with linking unseen entities in a new domain.

A Generate-and-Rank Framework with Semantic Type Regularization for Biomedical Concept Normalization

This paper proposes an architecture consisting of a candidate generator and a list-wise ranker based on BERT that achieves state-of-the-art performance on multiple datasets and enhances it with a semantic type regularizer that allows the model to incorporate semantic type information from the ontology during training.

Parallel sequence tagging for concept recognition

This work proposes a parallel architecture, where both NER and NEN are modeled as a sequence-labeling task, operating directly on the source text, and shows that the strengths of the two classifiers can be combined in a fruitful way.

TaggerOne: joint named entity recognition and normalization with semi-Markov Models

This work proposes the first machine learning model for joint NER and normalization during both training and prediction, which is trainable for arbitrary entity types and consists of a semi-Markov structured linear classifier, with a rich feature approach for N ER and supervised semantic indexing for normalization.