2019/10/08 by Usama Yaseen, Pankaj Gupta, Yaseen, Usama +3
Biochemistry, Genetics and Molecular Biology · Computer Science · #Biomedical Text Mining and Ontologies #Computation and Language (cs.CL) #FOS: Computer and information sciences #Natural Language Processing Techniques #Topic Modeling
paper · pdf · doi:10.48550/arxiv.1910.03385
openalex publication_date 2019/10/08 · openalex created_date 2022/07/28 · openalex updated_date 2026/07/28
Named Entity Recognition (NER) and Relation Extraction (RE) are essential\ntools in distilling knowledge from biomedical literature. This paper presents\nour findings from participating in BioNLP Shared Tasks 2019. We addressed Named\nEntity Recognition including nested entities extraction, Entity Normalization\nand Relation Extraction. Our proposed approach of Named Entities can be\ngeneralized to different languages and we have shown it's effectiveness for\nEnglish and Spanish text. We investigated linguistic features, hybrid loss\nincluding ranking and Conditional Random Fields (CRF), multi-task objective and\ntoken-level ensembling strategy to improve NER. We employed dictionary based\nfuzzy and semantic search to perform Entity Normalization. Finally, our RE\nsystem employed Support Vector Machine (SVM) with linguistic features.\n Our NER submission (team:MIC-CIS) ranked first in BB-2019 norm+NER task with\nstandard error rate (SER) of 0.7159 and showed competitive performance on\nPharmaCo NER task with F1-score of 0.8662. Our RE system ranked first in the\nSeeDev-binary Relation Extraction Task with F1-score of 0.3738.\n