Research IT application approved on reader responses to translated literature
Together with prof. dr. Haidee Kotze (Translation Studies) the DHLab has started a new research project on exploring reader responses to translated literature using big data. The IT innovation project has wide application, as it is of relevance to all disciplines where scholars deal with multilingual and/or translated data.
Little is known about how real “non-professional” readers, across different contexts and languages, perceive, respond to, and assess translated literary texts. This project focuses on user-generated reviews of literary texts translated in multiple language pairs to study readers’ perceptions of translations. By bringing the technique of word embeddings to the field of translation studies, we intend to simultaneously explore their potential in doing multilingual research as well as their added value in performing sentiment analysis on social data (in the form of user reviews).
Traditionally, research on the reception of literary texts has made use of reviews produced by a small pool of “expert” reviewers. Our innovation will allow for the large-scale quantitative analysis of translation reception by real readers across genres and language pairs. It will consolidate UU’s established tradition of literary translation criticism, while expanding it into the domain of digital humanities and AI. Beyond this innovation within the discipline of translation studies, the IT innovation will contribute to the debates around sentiment analysis on online user reviews based on word embeddings (Bansal & Srivastava 2016, Giatsoglou et al. 2017) and the impact of multilingual word vector models for textual research that crosses language barriers.