يتطلب فهم اللغة الطبيعية الحس السليم، وهو جانب واحد منها هو القدرة على تمييز معقول الأحداث.في حين أن نماذج التوزيع --- أحدث نماذج لغة محول مؤخرا --- - - أظهرت تحسينات في حالة قدر نفواد الحدث، فإن أدائها لا يزال أقل من البشر.في هذا العمل، نظهر أن نماذج المعقول القائم على المحولات لا تتفق عليها بشكل ملحوظ عبر الفصول المفاهيمية للتسلسل الهرمي المعجمي، مما يستنتج أن الشخص يتنفس "من المعقول بينما يتنفس طبيب الأسنان" ليس كذلك، على سبيل المثال.نجد أن هذا التناقض يستمر حتى عندما يتم حقن الطراجات بهدوء مع المعرفة المعجمية، ونقدم طريقة بسيطة ما بعد الهوك لإجبار الاتساق النموذجي الذي يحسن الارتباط مع أحكام الصفقات البشرية.
Understanding natural language requires common sense, one aspect of which is the ability to discern the plausibility of events. While distributional models---most recently pre-trained, Transformer language models---have demonstrated improvements in modeling event plausibility, their performance still falls short of humans'. In this work, we show that Transformer-based plausibility models are markedly inconsistent across the conceptual classes of a lexical hierarchy, inferring that a person breathing'' is plausible while a dentist breathing'' is not, for example. We find this inconsistency persists even when models are softly injected with lexical knowledge, and we present a simple post-hoc method of forcing model consistency that improves correlation with human plausibility judgements.
References used
https://aclanthology.org/
We present a system for learning generalized, stereotypical patterns of events---or schemas''---from natural language stories, and applying them to make predictions about other stories. Our schemas are represented with Episodic Logic, a logical form
Storytelling, whether via fables, news reports, documentaries, or memoirs, can be thought of as the communication of interesting and related events that, taken together, form a concrete process. It is desirable to extract the event chains that repres
Prior research has explored the ability of computational models to predict a word semantic fit with a given predicate. While much work has been devoted to modeling the typicality relation between verbs and arguments in isolation, in this paper we tak
Universal Conceptual Cognitive Annotation (UCCA) is a semantic annotation scheme that organizes texts into coarse predicate-argument structure, offering broad coverage of semantic phenomena. At the same time, there is still need for a finer-grained t
We study the problem of Event Causality Identification (ECI) to detect causal relation between event mention pairs in text. Although deep learning models have recently shown state-of-the-art performance for ECI, they are limited to the intra-sentence