[Back]

C. Ivanovic, A. Frank:
"Corpus-based Research in Computational Comparative Literature";
Talk: Workshop on Corpus-Based Research in the Humanities (CRH), Warschau; 2015-12-10; in: "Proceedings of the Workshop on Corpus-Based Research in the Humanities (CRH)", F. Mambrini, M. Passarotti, C. Sporleder (ed.); Polish Academy of Sciences, (2015), ISBN: 978-83-63159-19-1; 69 - 78.

Comparative literature investigates texts simultaneously in multiple languages. It is by definition conducting corpus-based research (each study is based on a quantitatively discernible number of texts chosen by single sets of qualities), in practice however it is facing diverse limitations which restrict the value of generalized statements. We propose to extend well approved methods of corpus-based research to computational comparative literature in order to overcome those limitations. The core is the construction of a large corpus of annotated literary texts with a related set of computational analysis methods. The aspects of the proposed technical solution make it even suited for the comparison of multi-lingual text corpora, as the same methods can be applied to all texts irrespective of language. Computational comparative literature has a supportive function for established approaches in comparative literature. It allows for systematic evaluation of a large number of texts that goes beyond established questions.

Comparative literature investigates texts simultaneously in multiple languages. It is by definition conducting corpus-based research (each study is based on a quantitatively discernible number of texts chosen by single sets of qualities), in practice however it is facing diverse limitations which restrict the value of generalized statements. We propose to extend well approved methods of corpus-based research to computational comparative literature in order to overcome those limitations. The core is the construction of a large corpus of annotated literary texts with a related set of computational analysis methods. The aspects of the proposed technical solution make it even suited for the comparison of multi-lingual text corpora, as the same methods can be applied to all texts irrespective of language. Computational comparative literature has a supportive function for established approaches in comparative literature. It allows for systematic evaluation of a large number of texts that goes beyond established questions.

Corpus-based Redsearch, Comparative Literature, Spatial Reference, Temporal Reference, Cluster Analysis