<![CDATA[This study presents a systematic literature review (SLR) of Translation Memory (TM) mechanisms in technology-based translation processes, explicitly following the PRISMA framework for identification, screening, eligibility, and inclusion of studies. A total of 26 journal articles and conference papers published between 2020 and 2025 were analyzed. Findings reveal that TM operates through core mechanisms, including segmentation, storage of translation units, matching, retrieval, exact matches, fuzzy matches, and similarity measurement, which improve translation consistency, efficiency, and productivity, especially in repetitive and specialized tasks. The study also identifies limitations of TM, including dependence on database quality, reduced contextual flexibility, and challenges in capturing semantic variation with conventional matching methods. Furthermore, recent studies show that TM has evolved beyond traditional CAT tools and is increasingly integrated into neural machine translation and retrieval-augmented systems. This review contributes to the field by synthesizing current research trends, highlighting gaps in the literature, and providing recommendations for future studies to advance the integration and application of TM in modern translation workflows.]]>
Copyrights © 2026
