Detection of Domain-Specific Hypernyms Using Named Entity Recognition and Extraction

• Published in: 2024 IEEE International Conference on Automation, Quality and Testing, Robotics (AQTR) pp. 1 - 5
• Main Authors: Prata, D.L., Mocan, C.M., Nandra, C.I., Chifu, E.S.
• Format: Conference Proceeding
• Language: English
• Published: IEEE 16-05-2024
• Subjects: Accuracy; Automated learning; Distributional models; Hypernym detection; Machine learning; Natural Language Processing; Semantic segmentation; Semantics; SemEval-2018; Solids; Tagging; Taxonomy learning; Time-frequency analysis; Vocabulary
• ISBN: 9798350361919
• ISSN: 1844-7872
• DOI: 10.1109/AQTR61889.2024.10554157

This paper explores the task of hypernym detection in the medical domain, leveraging natural language processing techniques and machine learning. We address this challenge by presenting a solution to the SemEval-2018 Task 9: Hypernym Discovery, focusing on identifying hypernyms for specific terms within a given search space. Our approach involves integrating pattern-based and distributional methods to achieve wider coverage while maintaining accuracy. We illustrate the process of hypernym detection as a sentence segmentation problem, where each word is labeled based on its semantic category. Through empirical evaluation, we demonstrate the effectiveness of our proposed solution, highlighting its performance metrics and execution times.