Select Page

Using Computers to Read Between the Lines

For Paul Felt, the combination of computer science and linguistics spells passion. With an undergraduate degree in English under his belt, Felt is working on adding a PhD in computer science to his academic tool belt.

Felt won a Nibley Fellowship from the Neal A. Maxwell Institute for Religious Scholarship for his research in computer annotations of ancient languages. The Nibley Fellowship is awarded to students who study historical—particularly religious—texts.

“Usually people who are applying for this fellowship are associated with ancient language scholars, not computer science,” Felt said. His advisor, Eric Ringger, helped him discover the fellowship.

“Eric Ringger is continually on the lookout for funding opportunities for his students, and he encouraged me to apply for the Nibley fellowship,” Felt said. Ringger also set up collaborative opportunities for Felt with Kristian Heal from the Maxwell Institute.

The two collaborated to study Syriac, a specific dialect of ancient Aramaic used by Christians in the Middle East. They used natural language processing (NLP) to study Syriac texts from the time of Christ. NLP applies machine-learning techniques to extract meaningful information and annotate it for organizational purposes.

Annotation through NLP can apply to a variety of fields. For example, email servers use NLP to sort junk emails by determining which words are typically found in a junk email and sorting them into folders.

Linguistically, Heal and Felt use NLP to label the different parts of speech in the Syriac language.

“[On] the linguistics side, we’re researching ancient languages for interest in the [Syriac] language,” Felt said.

From a cultural standpoint, this interest in the language could unravel some writings and details of Syriac culture. Many Syriac writings haven’t been well studied due to a lack of resources, but Heal and Felt’s research is contributing to digitizing a body of manuscripts and data while labeling approximately 10 million words of text.

In this sense, Felt’s research contributes not only to findings related to annotation and NLP, but also sheds new light on religious studies.

By Alysa Kleinman Posted on