Alert button
Picture for Sharid Loáiciga

Sharid Loáiciga

Alert button

Unsupervised Discovery of Unaccusative and Unergative Verbs

Nov 01, 2021
Sharid Loáiciga, Luca Bevacqua, Christian Hardmeier

Figure 1 for Unsupervised Discovery of Unaccusative and Unergative Verbs
Figure 2 for Unsupervised Discovery of Unaccusative and Unergative Verbs

We present an unsupervised method to detect English unergative and unaccusative verbs. These categories allow us to identify verbs participating in the causative-inchoative alternation without knowing the semantic roles of the verb. The method is based on the generation of intransitive sentence variants of candidate verbs and probing a language model. We obtained results on par with similar approaches, with the added benefit of not relying on annotated resources.

Viaarxiv icon

Is Incoherence Surprising? Targeted Evaluation of Coherence Prediction from Language Models

May 07, 2021
Anne Beyer, Sharid Loáiciga, David Schlangen

Figure 1 for Is Incoherence Surprising? Targeted Evaluation of Coherence Prediction from Language Models
Figure 2 for Is Incoherence Surprising? Targeted Evaluation of Coherence Prediction from Language Models
Figure 3 for Is Incoherence Surprising? Targeted Evaluation of Coherence Prediction from Language Models
Figure 4 for Is Incoherence Surprising? Targeted Evaluation of Coherence Prediction from Language Models

Coherent discourse is distinguished from a mere collection of utterances by the satisfaction of a diverse set of constraints, for example choice of expression, logical relation between denoted events, and implicit compatibility with world-knowledge. Do neural language models encode such constraints? We design an extendable set of test suites addressing different aspects of discourse and dialogue coherence. Unlike most previous coherence evaluation studies, we address specific linguistic devices beyond sentence order perturbations, allowing for a more fine-grained analysis of what constitutes coherence and what neural models trained on a language modelling objective do encode. Extending the targeted evaluation paradigm for neural language models (Marvin and Linzen, 2018) to phenomena beyond syntax, we show that this paradigm is equally suited to evaluate linguistic qualities that contribute to the notion of coherence.

* Accepted as long paper at NAACL 2021 
Viaarxiv icon