SciCo: Hierarchical Cross-Document Coreference for Scientific Concepts
This addresses a gap in natural language understanding for scientific and technical domains, where incremental improvements are made by focusing on hierarchical structures in coreference.
The paper tackles the problem of cross-document coreference resolution for complex scientific concepts, which involve hierarchical relationships and ambiguity, by introducing a new task called Hierarchical CDCR and creating SciCo, a dataset 3X larger than existing resources like ECB+.
Determining coreference of concept mentions across multiple documents is a fundamental task in natural language understanding. Previous work on cross-document coreference resolution (CDCR) typically considers mentions of events in the news, which seldom involve abstract technical concepts that are prevalent in science and technology. These complex concepts take diverse or ambiguous forms and have many hierarchical levels of granularity (e.g., tasks and subtasks), posing challenges for CDCR. We present a new task of Hierarchical CDCR (H-CDCR) with the goal of jointly inferring coreference clusters and hierarchy between them. We create SciCo, an expert-annotated dataset for H-CDCR in scientific papers, 3X larger than the prominent ECB+ resource. We study strong baseline models that we customize for H-CDCR, and highlight challenges for future work.