In a recent presentation at the Department of Old Russian Literature of The Pushkinsky House (Institute of Russian Literature), we presented our project explored a novel vector-based approach to the stemma codicum of “Моление Даниила Заточника”. This project aimed to map the genealogical relationships of its witnesses using computational methods.

Elena Ripinskaya, a PhD candidate at the Institute of Russian Literature, provided the foundational material and the project’s aim. She compiled a comprehensive table documenting the variations in each line across 11 manuscripts, setting the stage for in-depth textual analysis.

My contribution was centered around the application of advanced computational techniques. I employed the word2vec algorithm to transform the manuscripts into semantic vector representations. Subsequently, I utilized KMeans clustering to systematically categorize these manuscripts. This methodology was pivotal in revealing distinct groups based on their textual similarities, thus bringing a new perspective to traditional stemmatological methods.

Sources’ description