|
|
|
News
 About me
I am Markos Mylonakis and I'm currently in my final year as a PhD
student at the University
of Amsterdam. I work with the Language
and Computation group
of the ILLC
(Institute for Logic, Language and Computation). My supervisor is Khalil Sima'an.
My research is on
natural language processing, using a combination of
statistical and machine
learning methods. I am particularly interested in statistical
machine translation and how to improve over state-of-the-art
performance by addressing fundamental problems.
The
last few years I have been progressively working towards learning
machine translation's latent structure, i.e. deciphering how
translation works as a recursive process.
I like to explore
how to learn translation models that take advantage of any possibly
useful information which might aid in translating like linguistic
annotations, while not becoming constrained by using them as a sole
vehicle to explain the translation process.
Furthermore, I am
interested in how to balance memorising with generalising, for example
by finding out when translation constructions should be better
memorised as whole units and when they should be left to be constructed
by building them from smaller translation blocks, focusing then on
learning how to combine these together.
You may find more information about ways to contact me, my publications
and released software by using the links on the left. |
|
| |
Interesting Links
Photos
Opinion & Analysis
|
|
PhD defence date set!