Dynamic Named Entity Recognition Article Swipe
YOU?
·
· 2023
· Open Access
·
· DOI: https://doi.org/10.1145/3555776.3577603
· OA: W4321524613
Named Entity Recognition (NER) is a challenging and widely studied task that\ninvolves detecting and typing entities in text. So far,NER still approaches\nentity typing as a task of classification into universal classes (e.g. date,\nperson, or location). Recent advances innatural language processing focus on\narchitectures of increasing complexity that may lead to overfitting and\nmemorization, and thus, underuse of context. Our work targets situations where\nthe type of entities depends on the context and cannot be solved solely by\nmemorization. We hence introduce a new task: Dynamic Named Entity Recognition\n(DNER), providing a framework to better evaluate the ability of algorithms to\nextract entities by exploiting the context. The DNER benchmark is based on two\ndatasets, DNER-RotoWire and DNER-IMDb. We evaluate baseline models and present\nexperiments reflecting issues and research axes related to this novel task.\n
