Boshko Koloski

L39, Dept. of Knowledge Technologies

39 Jamova cest

Ljubljana, Slovenia 1000

I’m a Ph.D. student at the Jožef Stefan International Postgraduate School. My research is focused on the intersection of Natural Language Processing and Learning on Graphs. I am keen on exploring how machine learning models perform in low-resource settings for less-represented languages, especially interested in transfer-learning within cross-lingual applications. I find the process of transferring acquired knowledge fascinating, and I equally enjoy the act of teaching.

I am Macedonian, born in Bitola, Macedonia, on 8th January 1999. I finished my elementary (OU. Todor Angelevski) and secondary school (SOU. Josip Broz Tito, natural-mathematics major) in my hometown. In 2017, I moved to Ljubljana, Slovenia, to pursue my further studies. I got my undergraduate degree from the Faculty of Computer and Information Science of the University of Ljubljana.

I love talking about science, beer, indie music bands, and video games.

selected publications

ICLR

Inducing Document Representations from Graphs: A Blueprint

Boshko Koloski, Marko Pranjic, Nada Lavrac, and 2 more authors

2023

Bib

@article{koloski2023inducing,
  title = {Inducing Document Representations from Graphs: A Blueprint},
  author = {Koloski, Boshko and Pranjic, Marko and Lavrac, Nada and {\v{S}}krlj, Bla{\v{z}} and Pollak, Senja},
  year = {2023}
}

Neurocomputing
Knowledge graph informed fake news classification via heterogeneous representation ensembles

Boshko Koloski, Timen Stepišnik Perdih, Marko Robnik-Šikonja, and 2 more authors

Neurocomputing, 2022

Abs Bib

Increasing amounts of freely available data both in textual and relational form offers exploration of richer document representations, potentially improving the model performance and robustness. An emerging problem in the modern era is fake news detection—many easily available pieces of information are not necessarily factually correct, and can lead to wrong conclusions or are used for manipulation. In this work we explore how different document representations, ranging from simple symbolic bag-of-words, to contextual, neural language model-based ones can be used for efficient fake news identification. One of the key contributions is a set of novel document representation learning methods based solely on knowledge graphs, i.e., extensive collections of (grounded) subject-predicate-object triplets. We demonstrate that knowledge graph-based representations already achieve competitive performance to conventionally accepted representation learners. Furthermore, when combined with existing, contextual representations, knowledge graph-based document representations can achieve state-of-the-art performance. To our knowledge this is the first larger-scale evaluation of how knowledge graph-based representations can be systematically incorporated into the process of fake news classification.
@article{KOLOSKI2022208, title = {Knowledge graph informed fake news classification via heterogeneous representation ensembles}, journal = {Neurocomputing}, volume = {496}, pages = {208-226}, year = {2022}, issn = {0925-2312}, doi = {https://doi.org/10.1016/j.neucom.2022.01.096}, url = {https://www.sciencedirect.com/science/article/pii/S0925231222001199}, author = {Koloski, Boshko and {Stepišnik Perdih}, Timen and Robnik-Šikonja, Marko and Pollak, Senja and Škrlj, Blaž}, keywords = {Fake news detection, Knowledge graphs, Text representation, Representation learning, Neuro-symbolic learning} }
LREC

Out of Thin Air: Is Zero-Shot Cross-Lingual Keyword Detection Better Than Unsupervised?

Boshko Koloski, Senja Pollak, Blaž Škrlj, and 1 more author

In Proceedings of the Thirteenth Language Resources and Evaluation Conference, Jun 2022

Abs

Keyword extraction is the task of retrieving words that are essential to the content of a given document. Researchers proposed various approaches to tackle this problem. At the top-most level, approaches are divided into ones that require training - supervised and ones that do not - unsupervised. In this study, we are interested in settings, where for a language under investigation, no training data is available. More specifically, we explore whether pretrained multilingual language models can be employed for zero-shot cross-lingual keyword extraction on low-resource languages with limited or no available labeled training data and whether they outperform state-of-the-art unsupervised keyword extractors. The comparison is conducted on six news article datasets covering two high-resource languages, English and Russian, and four low-resource languages, Croatian, Estonian, Latvian, and Slovenian. We find that the pretrained models fine-tuned on a multilingual corpus covering languages that do not appear in the test set (i.e. in a zero-shot setting), consistently outscore unsupervised models in all six languages.