SUPR
Topic modeling for ancient novels
Dnr:

NAISS 2024/22-361

Type:

NAISS Small Compute

Principal Investigator:

Emelie Hallenberg

Affiliation:

Uppsala universitet

Start Date:

2024-03-13

End Date:

2025-04-01

Primary Classification:

60204: Specific Literatures

Webpage:

Allocation

Abstract

Topic modeling for ancient languages is still underdeveloped. In my Ph.D. project in Ancient Greek, part of the national research school DigPhil (funded by Vetenskapsrådet), I will analyze Ancient Greek novels using various NLP methods, including topic modeling. The aim of this study is, first, to evaluate the available approaches and language models for topic in Ancient Greek novels, and second, to conduct a study of the frequency and distribution of prominent themes and motifs in these texts. Access to GPUs will allow fine-tuning and evaluation of larger models for the task.