Design and build the prototype for a model and application for translating in conversation with machines using new and emerging technologies of neural machine translation (NMT) and natural language processing (NLP). Specifically, we propose to build and train an NMT model for engaging in translation of Hebrew biblical literature, and then to build a web application that enables users (students, fellow scholars, and the general public) to explore different translational possibilities and create their own translations in conversation with the NMT model. The initial prototype of the model and application will be based on the first creation story in the Hebrew Scriptures (Genesis 1:1-2:4a). Although the model and application will be based on translating Hebrew biblical texts, they will be adaptable to other texts in other languages for which there exist multiple translations. The prototype will also be scalable, with the next step being full implementation to incorporate the entire canon of Hebrew scriptures.
See samples of data and code here on Open Science Framework.
Project team includes Justin Barber, Timothy Beal, and Michael Hemenway.