My name is Mun Thye, and I am a graduate of Carnegie Mellon University's School of Computer Science. I graduated in May 2009 with a Bachelor's of Science in Computer Science with University Honours with a minor in Mathematical Sciences.
Contrary to what it might seem, my first name is Mun Thye, and is my preferred form of address; the compound form of Mun-Thye may appear in publications but that exists solely to reduce confusion for cultures unused to having more than one word for a first name. The official English pronounciation of my name is [mɐn tʰaːɪ].
My focus is on machine learning, and have performed research with Prof Tom Mitchell and Prof Estevam Hruschka Jr in Fall 2008 on how best to train better classifiers that can differentiate noun phrases based on the context in which they appear in. The final report can be downloaded here.
In slightly more detail, my research interest is in the learning of various pieces of knowledge from text, be it from traditional sources or that of the Internet. Unlike traditional Natural Language Processing (NLP) and Information Extraction (IE) approaches, I am focused on using both the text and the meta-data that is often associated with the text to achieve better classifying and information extraction results than using the text alone. Eventually, I would like to apply machine learning to computer program source codes to learn abstract concepts with an application of automatically fine-tuning the algorithm that is presented in the program.
Apart from my primary research in machine learning in the text domain, my general computer science interests include systems hacking, cryptography and steganography, algorithms, as well as machine learning over various disciplines and data sources.
I am currently based in Singapore, working in the Data Mining Department of the Institute of Infocomm Research under supervision of Dr Ng See-Kiong. Other collaborators that I am working with include Dr Sujoy Roy, Dr Hady Lauw and Prof Kevin Chang. The projects I am involved in deal with on-demand information extraction using machine learning techniques, a happy marriage between the highly theoretical parts of machine learning and the applied aspects of data mining.
The most updated version of my resume can be obtained through direct email. I am happy to furnish other forms of documentation with regards to my academic record, but again, only upon request.
Even though I am currently based in a fixed location, I am hardly near a fixed telephone---any inconvenience is regretted. Please revert to using email. Thank you for your understanding.
Last updated on 2010 Jun 09.