Language Technologies in Humanities: Computational Semantic Analysis in Folkloristics

The paper discusses computational methods for natural language processing (NLP) and possibilities they offer to folkloristics. As folkloristic materials are very challenging for NLP, due to their specific semanticsyntactic structure, inherent dialectical diversity and strong intertextuality, a robust NLP method is needed that can account for topical distribution, detect general heterogeneity, and context. The focus of this paper is on computational semantic analysis (such as word-sense disambiguation, topic recognition) and its ability to uncover latent semantic structure of folkloristic corpora.