Sejong Korean Corpora in the Making

The 21st Century Sejong Project is a comprehensive project aiming to build various kinds of language resources including Korean corpora, comparable to BNC (Aston & Burnard, 1998), and Korean electronic dictionaries. The project was conceived of in 1997 and started in 1998 as a 10-year long-term project. By 2003, we completed 6 years of our work. The Sejong Corpora are a collection of raw corpora of modern Korean (written and spoken), North Korean, Korean used abroad, old Korean, and oral folklore literature. They also include parallel corpora consisting of Korean and other languages such as English and Japanese. Among these, a morph tagged corpus is a central part. In the process of compiling these corpora we followed suggestion from Text Encoding Initiative (TEI, Sperberg-McQueen & Burnard, 1994) to a certain degree. By 2003, we compiled a modern Korean raw corpus of 57 million words. We have additional 75 million words of already existing electronic texts which were processed and standardized in the first year of the Sejong project. These raw texts are mostly written Korean. We have relatively small amount, around 3 million, of spoken words. The morph tagged corpus is morphologically analyzed written Korean, around 10 million words by the end of 2003. The morph sense tagged corpus, which is the result of disambiguation of morphs, has 5.5 million words. From 2002 we started to build a treebank, i.e. syntactically analyzed Korean sentences on the basis of simple phrase structure grammar rules. Currently, we only have 0.15 million words being part of syntactic trees. Written corpora, i.e. a raw corpus of modern Korean, a morph tagged corpus, a morph sense tagged corpus, and a treebank, have been compiled at Center for Electronic Texts of Korea University. The following table is a summary.