Computing the Dharma: a natural language processing infrastructure to explore word meanings in Buddhist Sanskrit literature

FAIN: HAA-277246-21

Mangalam Centers (Berkeley, CA 94704-1418)
Ligeia Lugli (Project Director: June 2020 to present)
Senja Pollack (Co Project Director: December 2020 to present)

Research into the application of natural language processing techniques to study the evolution of language in Buddhist Sanskrit texts.

This application is for a Level II DHAG. The project has two objectives: 1) to advance research in Indian Buddhism by developing semi-automated methods to study the vocabulary of Buddhist Sanskrit texts; and 2) to contribute to the Digital Humanities by refining computational methods that leverage representations of words as numerical vectors. These vector representations of language, called "word embedding models," have found wide application in industry and are gaining traction in Humanities research. Due to their technical complexity, however, the full potential of cutting-edge word embedding techniques is rarely deployed in the Humanities, and best practices for reliably applying them to the study of historical texts are yet to be drafted. This project brings together Natural Language Processing experts and Buddhist Sanskrit scholars to devise and test new methods for harnessing the power of latest-generation word embedding techniques for historical textual scholarship.