20 September 2019 - Reut Tsafraty: Seminar
Title: Finding NEMO in the Deep: Neural models for Morphology in the Deep learning Era
Morphologically Rich Languages (MRLs) are long known to pose unique challenges to language technology and to the standard NLP pipeline. Does the introduction of Deep Learning (DL) and Neural Network (NN) models places us in a better position towards solving them? In this talk I aim to answer this question. I survey key challenges in parsing MRLs, discuss pre-neural approaches towards their resolution, and assess the plausiblity of their neural counterparts. I show preliminary results of experiments probing the ability of NN models to solve NLP tasks in a morphologically-aware way. Based on these experiments, I argue that NN models (and the abundance of unsupervised data) do not, in and of themselves, solve the key challenges in MRLs. Instead, I present a new set of research questions regarding architectural and representational choices in NN modeling, of which resolution can potentially push NLP performance on MRLs to a new level.
Reut Tsarfaty is an Assistant Professor at the Open University of Israel (soon to be Associate Professor at Bar-llan university), heading the Open Natural Language Processing research lab (The ONLP Lab). Reut holds a BSc. from the Technion and MSc./PhD. from the Institute for Logic Language and Computation (ILLC) at the University of Amsterdam. She also held postdoctoral fellowships at Uppsala University in Sweden and at the Weizmann Institute in Israel. Her research focuses on parsing, broadly interpreted to cover morphological, syntactic and semantic phenomena, of typologically different languages. Applications she has worked on include (but are not limited to) natural language programming, natural language navigation, automated essay scoring, analysis and generation of social media content, and more. The research at the ONLP Lab is funded by an ERC-Starting-Grant #677352 and ISF grant #1739/26.