Joakim Nivre, Institutionen för lingvistik och filologi, Uppsala universitet
In this talk I will begin by giving a brief overview of the language technology
group at Uppsala University and some of our ongoing projects. I will then zoom in on three of these projects and present some of our recent work. The first is the Swedish Treebank project, where we are trying to create a largish treebank for Swedish by recycling two existing corpora: Talbanken and SUC. The second is an attempt to use machine learning to model human eye movement during reading. The third is a project aiming to develop efficient and accurate algorithms for dependency-based syntactic parsing, focusing in particular on the analysis of so-called non-projective dependency structures.