Abstract
We present a number of experiments on parsing the Ancient Greek Dependency\r\nTreebank (AGDT), i.e. the largest syntactically annotated corpus of\r\nAncient Greek currently available (350k words ca). Although the AGDT is\r\nrather unbalanced and far from being representative of all genres and periods\r\nof Ancient Greek, no attempt has been made so far to perform automatic\r\ndependency parsing of Ancient Greek texts. By testing and evaluating one\r\nprobabilistic dependency parser (MaltParser), we focus on how to improve\r\nthe parsing accuracy and how to customize a feature model that fits the distinctive\r\nproperties of Ancient Greek syntax. Also, we prove the impact of\r\ngenre and author diversity on parsing performances.
Lingua originale | English |
---|---|
Titolo della pubblicazione ospite | Proceedings of the Eleventh International Workshop on Treebanks and Linguistic Theories (TLT11). 30 November – 1 December 2012, Lisbon, Portugal |
Editore | Edições Colibri |
Pagine | 133-144 |
Numero di pagine | 12 |
ISBN (stampa) | 978-989-689-274-6 |
Stato di pubblicazione | Pubblicato - 2012 |
Keywords
- Linguistic theories
- Treebanks