About the journal

• Home page
• Contact information
• Aims and scope
• Indexing information
• Editorial policies
• ComSIS consortium
• Journal boards
• Managing board

For authors

• Information for contributors
• Article submission
• Accepted papers
• Copyright transfer form
• Download section

For readers

• Forthcoming articles
• Current issue
• Archive

For reviewers

• Review submissions
• Review submissions OJS

News

• Journal's Facebook page
• Special issue proposal
• Calls for special issues

A Pilot Study of Multi-Method Evaluation of Machine Translation in Macedonian

Jana Kuzmanova¹   , Katerina Zdravkova²   

1. Faculty of Computer Science and Engineering
   jana.kuzmanova@finki.ukim.mk
2. Faculty of Computer Science and Engineering
   katerina.zdravkova@finki.ukim.mk (corresponding author)

Abstract

This pilot study offers a linguistic evaluation of six machine translation systems: GPT-4o, GPT-5, Gemini 2.5 Flash, Google Translate, Microsoft Translator, and NLLB-600M applied to the translation of a short excerpt of Orwell’s “1984” into Macedonian. The analysis consisted of three interconnected experiments: manual annotation of translation errors and comparison with human output, evaluation using eight popular MT metrics, and sentence-level similarity analysis via cosine similarity, Jaccard similarity, and Levenshtein distance. Manual annotation revealed that stylistic errors (48.47%) and linguistic errors (34.54%) were the most common. The LLMs outperformed other systems, particularly GPT-5, while NLLB-600M performed poorly, often introducing incomprehensible sentences or non-existent words. Metrics-based evaluation showed that lexical metrics sometimes penalized fluent and accurate translations that deviated from the reference. Sentence similarity analysis confirmed that accurate translations were more consistent, while wrong–wrong sentence pairs were more divergent, especially in Levenshtein scores. The findings underscore the importance of combining manual and metric-based evaluation to fully understand MT quality, particularly in low-resource settings.

Key words

Machine Translation, Manual Evaluation and Annotation, Linguistic Similarity, Low Resource Language

Digital Object Identifier (DOI)

https://doi.org/10.2298/CSIS250910201K

Publication information

Volume 23, Issue 2 (April 2026)
Year of Publication: 2026
ISSN: 2406-1018 (Online)
Publisher: ComSIS Consortium

Full text

Available in PDF
Portable Document Format

How to cite

Kuzmanova, J., Zdravkova, K.: A Pilot Study of Multi-Method Evaluation of Machine Translation in Macedonian. Computer Science and Information Systems, 23(2), 827–860 (2026). https://doi.org/10.2298/CSIS250910201K