Development of a benchmark corpus to support the automatic extraction of drug-related adverse effects from medical case reports

By Harsha Gurulingappa, Abdul Mateen Rajput, Angus Roberts, Juliane Fluck, Martin Hofmann-Apitius, Luca I. G. Toldo. 2012

493 citations recorded by the source corpus.

Subfields: Computational Theory and Mathematics, Molecular Biology, Toxicology, Natural Language Processing, Medical & Health AI

How researchers see this paper

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

In Computational Theory and Mathematics

In Molecular Biology

In Toxicology

In Natural Language Processing

In Medical & Health AI

Jie Tang

Citations in the corpus

FedNER: Medical Named Entity Recognition with Federated Learning (2020, methodology): “We experiment on three publicly available medical NER datasets, e.g., C ADEC (Karimi et al., 2015), ADE Corpus (Gurulingappa et al., 2012) and SMM4H (Weissenbacher et al., 2019).”
Assessing the state of the art in biomedical relation extraction: overview of the BioCreative V chemical-disease relation (CDR) task (2016, background): “A small number of test corpora were also developed, but they are limited in size and annotation scope (10, 11).”
The DDI corpus: An annotated corpus with pharmacological substances and drug-drug interactions (2013, background): “Other corpora such as ADE (Adverse Drug Effect) [22], EU-ADR (Exploring and Understanding Adverse Drug Reactions) [2] or ITI TXM (Tissue Expressions and Protein–Protein Interactions) [23] use a single entity type to annotate both drugs and chemicals, while the BioCaster [14] corpus distinguishes bet”
Portable Automatic Text Classification for Adverse Drug Reaction Detection via Multi-corpus Training (2014, methodology): “[6] used the publicly available ADE corpus [34] to perform binary classification using various classifiers, obtaining a maximum F-score of 0.77 for the ADR class.”
DeepStruct: Pretraining of Language Models for Structure Prediction (2022, background): “We also find the multi-task DEEPSTRUCT performs competitively with previous task-specific approaches on both ADE and NYT.”
BioCreative V CDR task corpus: a resource for chemical disease relation extraction (2016, background): “The ADE corpus consists of 2972 PubMed articles with sentence-level statements of 5776 adverse effects related to 5063 drugs (17).”
RAFT: A Real-World Few-Shot Text Classification Benchmark (2021, background): “The ADE corpus V2 [15] contains sentences from medical case reports annotated for relation to adverse drug effects.”
Extraction of potential adverse drug events from medical case reports (2012, methodology): “[11], we adopted the standard established by the CLEF framework for the annotation workflow [10] however we reshaped the annotation schema by using only two of the original entities (condition, drug) and extended it with a third one (dosage).”
Span-based Joint Entity and Relation Extraction with Transformer Pre-training (2019, background): “• ADE: The ADE dataset (Gurulingappa et al. 2012) consists of 4, 272 sentences and 6, 821 relations extracted from medical reports that describe the adverse effects arising from drug use.”
FedNER: Privacy-preserving Medical Named Entity Recognition with Federated Learning (2020, methodology): “We randomly select different ratios of data for model training, and due to space limit we only show the results on ADE Corpus dataset in Figure 3.”
Communication-efficient federated learning via knowledge distillation (2021, unknown): “We conduct experiment on three public benchmark medical NER datasets, e.g., CADEC 32 , ADE 33 and SMM4H 34 .”
Joint entity recognition and relation extraction as a multi-head selection problem (2018, methodology): “Specifically, we apply our method to four datasets, namely ACE04 (news), Adverse Drug Events (ADE), Dutch Real Estate Classifieds (DREC) and CoNLL’04 (news).”
Improving chemical disease relation extraction with rich features and weakly labeled data (2016, methodology): “For each pair, useful information from NLP tools including part-of-speech taggers, full parsers, and dependency parsers were extracted as features [20, 56].”