Counts@IITK at SemEval-2021 Task 8: SciBERT Based Entity And Semantic Relation Extraction For Scientific Data
This work addresses entity and relation extraction for scientific data, but it is incremental as it applies an existing method (SciBERT with CRF) to a new dataset.
The paper tackled the MeasEval task for extracting quantities, attributes, and relations from scientific text, achieving fifth place overall with top rankings in specific subtasks like Quantity and Unit.
This paper presents the system for SemEval 2021 Task 8 (MeasEval). MeasEval is a novel span extraction, classification, and relation extraction task focused on finding quantities, attributes of these quantities, and additional information, including the related measured entities, properties, and measurement contexts. Our submitted system, which placed fifth (team rank) on the leaderboard, consisted of SciBERT with [CLS] token embedding and CRF layer on top. We were also placed first in Quantity (tied) and Unit subtasks, second in MeasuredEntity, Modifier and Qualifies subtasks, and third in Qualifier subtask.