Ümit Mersinli


2012

pdf bib
Construction of the Turkish National Corpus (TNC)
Yeşim Aksan | Mustafa Aksan | Ahmet Koltuksuz | Taner Sezer | Ümit Mersinli | Umut Ufuk Demirhan | Hakan Yılmazer | Gülsüm Atasoy | Seda Öz | İpek Yıldız | Özlem Kurtoğlu
Proceedings of the Eighth International Conference on Language Resources and Evaluation (LREC'12)

This paper addresses theoretical and practical issues experienced in the construction of Turkish National Corpus (TNC). TNC is designed to be a balanced, large scale (50 million words) and general-purpose corpus for contemporary Turkish. It has benefited from previous practices and efforts for the construction of corpora. In this sense, TNC generally follows the framework of British National Corpus, yet necessary adjustments in corpus design of TNC are made whenever needed. All throughout the process, different types of open-source software are used for specific tasks, and the resulting corpus is a free resource for non-commercial use. This paper presents TNC's design features, web-based corpus management system, carefully planned workflow and its web-based user-friendly search interface.