Research

Andrei's research area covers scalable semi-supervised learning with graphs, and combination thereof with other learning techniques, applied to NLP problems, particularly on resource-poor languages. He works under the direction of Katrin Kirchhoff.

Publications

A. Alexandrescu and K. Kirchhoff, "Graph-Based Learning for Statistical Machine Translation", Proceedings of the North American Chapter of the Association for Computational Linguistics - Human Language Technologies (NAACL HLT) 2009 conference [PDF] [BibTeX].

A. Alexandrescu and K. Kirchhoff, "Graph-Based Learning for Phonetic Classification", Proceedings of The 2007 IEEE Automatic Speech Recognition and Understanding (ASRU) Workshop [PDF] [BibTeX]

A. Alexandrescu and K. Kirchhoff, "Data-Driven Graph Construction for Semi-Supervised Graph-Based Learning in NLP", Proceedings of HLT-NAACL 2007 [PDF] [BibTeX]

A. Alexandrescu and K. Kirchhoff, "Factored Neural Language Models", Proceedings of HLT-NAACL 2006 [PDF] [BibTeX]

A. Alexandrescu and K. Kirchhoff, "Factored Neural Language Models", UW EE Technical Report UWEETR-2006-0014 [HTML abstract] [PDF] [Gzipped PS] [BibTeX]

Doctoral Dissertation

Andrei's dissertation (a.k.a. thesis) is a tad unusual because it includes as-of-yet unpublished research (usually the doctoral dissertation is an expansion of the author's published research). Also, the text aims at being more readable than others, but it is unclear whether it succeeded at that. Doctoral dissertations are highly specialized and need to build extensive background, which inevitably ends up in referring to previous work that the putative reader would need to absorb in addition to the dissertation proper.

Below is a quick list of the contributions of the dissertation and other items of possible interest:

Read Andrei's dissertation: