Natural Language Processing, Social Text, Text Normalization
During my MS, I studied two different topics: Semantic Analysis and Text Normalization. I wrote my MS thesis on Text Normalization of Informal Text. I aimed to understand the dynamics of non-standard (Out of Vocabulary - OOV) words using contextual information within the phrase. Our model was based on a word association graph where the contextual relations between words are translated into directed edges of several degrees, as well as the lexical features. Our paper "A Graph-based Approach for Contextual Text Normalization" was published in Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP 2014).
Semantic analysis: I have been working on semantic classification of movie reviews. A small dataset of Turkish movie reviews crawled from beyazperde.com can be found here.
In the first year of my PhD studies, I took part in a Tübitak Funded project where we developed a Question Answering System called Hazircevap for the use of Turkish high school students. My main focus on this project was on Cross Lingual Information Retrieval (CLIR). I extended Hazircevap system such that it was able to find answers in resources in English and to translate them back to Turkish. Link for EBA resources.
Currently I am working as a part of a team of three with Erdem Yoruk and Deniz Yuret from Koc University on a research project: "Information extraction about social movements of the last decade in Brazil using newswire text as a part of a historical comparative sociological research project on the relationship between social welfare systems and movements in Brazil, Turkey and Korea". This project is funded by the European Commission’s Marie Curie Program. Our main goal is to extract event information from newswire text using Machine Learning techniques.
Department of Computer Engineering, Boğaziçi University,
34342 Bebek, Istanbul, Turkey
Phone: +90 212 359 45 23/24
Fax: +90 212 2872461
Connect with us
We're on Social Networks. Follow us & get in touch.