Event Date: Wednesday, 13 June, 2018, 11 a.m.
Location: Via Santa Maria, 36, Pisa, PI, Italia [2nd floor seminar room]
Speaker: Prof. Thierry Poibeau (LATTICE, CNRS, Paris)
Title: Multilingual Dependency Parsing for Low-Resource Languages
Abstract: I will present a method for dependency parsing using multilingual word embeddings. I will detail two main contributions. First, we propose a simple approach to building a bilingual dictionary and multilingual word embeddings for low-resource languages. Second, we show a model transfer parsing approach by using high-resource languages as a base model for parsing very low-resource languages. The multilingual approach outperforms the monolingual approach for resource-rich languages, but is especially useful for low resource languages. I will show some results for Finno-Ugric languages like North Saami and Komi. Joint work with KyungTae Lim and Niko Partanen (both at LATTICE, Paris)
Thierry Poibeau is a CNRS Director of Research and head of the LATTICE laboratory (Langues, Textes, Traitements informatiques et Cognition) since 2012. He is also an Affiliated Lecturer at the Department of Theoretical and Applied Linguistics (DTAL) of the University of Cambridge. He mainly works on Natural Language Processing (NLP) and linguistics, especially on the following topics: Information Extraction, Question Answering, Semantic Zoning, Knowledge Acquisition from text and Named Entity taggin