New Textual Corpora for Serbian Language Modeling
This provides new datasets for researchers working on Serbian language models, but it is incremental as it focuses on data collection rather than methodological innovation.
The paper tackles the lack of textual resources for Serbian language modeling by introducing new corpora, including a web corpus, a dissertation-based corpus, and a parallel translation corpus, and analyzes their uniqueness using stylometric methods.
This paper will present textual corpora for Serbian (and Serbo-Croatian), usable for the training of large language models and publicly available at one of the several notable online repositories. Each corpus will be classified using multiple methods and its characteristics will be detailed. Additionally, the paper will introduce three new corpora: a new umbrella web corpus of Serbo-Croatian, a new high-quality corpus based on the doctoral dissertations stored within National Repository of Doctoral Dissertations from all Universities in Serbia, and a parallel corpus of abstract translation from the same source. The uniqueness of both old and new corpora will be accessed via frequency-based stylometric methods, and the results will be briefly discussed.