NINJAL-UHM Linguistics Workshop Pre-Registration

First, I introduce Headless Relative Clauses and how they differ from related and better-known constructions like headed relative clauses and wh-interrogative clauses. Then, I touch on the main reasons why Headless Relative Clauses should be studied in general and across languages. Finally, I focus on the methodology that I have been developing for my cross-linguistic investigation of Headless Relative Clauses, in which I combine insights from both morpho-syntax and semantics. *This workshop is now full.

Entailment recognition (RTE) is a central task in natural language processing. The design of the task is simple: the system decides whether a conclusion (written in natural language) follows from a set of premises (also written in natural language). Due to an increasing interest in exploiting NLP resources in linguistic research and a growing awareness in NLP research that quality control in datasets is essential in improving the performance of implemented systems, both linguists and NLP researchers now recognize the need for carefully analyzing RTE datasets from a linguistic viewpoint. In this workshop,we'll take a look at the construction of a linguistically-oriented entailment recognition dataset, experience the challenges inherent to such a task, and discuss possible ways in which linguists can possibly contribute to this process. *This workshop is now full.