A special issue of Contributions to the Contemporary History has been published

A special issue of Contributions to the Contemporary History has been published, featuring research articles written as part of the project Large Language Models for Digital Humanities.

• Kaja Dobrovoljc presents a new version of the spoken Slovenian treebank, enriched with over 3,000 newly segmented utterances, emphasising the differences between spoken and written language and their influence on parser performance.

• Luka Terčon, Kaja Dobrovoljc, and Nikola Ljubešić outline the development and features of the CLASSLA-Stanza tool, which builds on the Stanza pipeline to enable accurate annotation of texts in South Slavic languages, including online resources and transcriptions of speech.

• Mojca Brglez, Veronika Bajt, Senja Pollak, Špela Rot, and Matej Martinc present a system for detecting semantic shifts in the Slovenian language, using the topic of migration as a case study to emphasise the significance of context and social influences in language use during specific periods.

• Špela Arhar Holdt, Magdalena Gapsa, Polona Gantar, and Iztok Kosem evaluate ChatGPT’s abilities in recognising synonyms, classifying them, and generating dictionary entries in Slovenian. They demonstrate that, despite certain limitations, it is a promising tool for digital lexicography.

Articles are available on this link: https://ojs.inz.si/pnz/index

To learn more about the project’s results and the broader possibilities of large language models in the digital humanities, we invite you to read the full special issue.