PulseAugur
LIVE 14:12:57
research · [2 sources] ·
0
research

Study finds machine translation distorts textual similarity in political manifestos

Researchers have investigated whether textual similarity is preserved after machine translation, using a corpus of political party platforms translated into English. They developed a framework to test the stability of similarity relationships across different embedding models and languages. Their analysis revealed that ten languages maintained translation invariance, while four showed detectable semantic distortion. AI

Summary written by gemini-2.5-flash-lite from 2 sources. How we write summaries →

IMPACT Provides a new methodology for evaluating the impact of machine translation on semantic similarity, applicable to downstream NLP tasks.

RANK_REASON Academic paper published on arXiv detailing a new framework for evaluating machine translation invariance.

Read on arXiv cs.CL →

COVERAGE [2]

  1. arXiv cs.CL TIER_1 · Daria Boratyn, Damian Brzyski, Albert Le\'sniak, Wojciech {\L}ukasik, Maciej Rapacz, Jan Rybicki, Wojciech S{\l}omczy\'nski, Dariusz Stolicki ·

    Is Textual Similarity Invariant under Machine Translation? Evidence Based on the Political Manifesto Corpus

    arXiv:2605.00618v1 Announce Type: new Abstract: We investigate the extent to which cosine similarity between paragraph embeddings is invariant under machine translation, using the Manifesto Corpus of over 2,800 political party platforms in 28 languages translated to English via t…

  2. arXiv cs.CL TIER_1 · Dariusz Stolicki ·

    Is Textual Similarity Invariant under Machine Translation? Evidence Based on the Political Manifesto Corpus

    We investigate the extent to which cosine similarity between paragraph embeddings is invariant under machine translation, using the Manifesto Corpus of over 2,800 political party platforms in 28 languages translated to English via the EU eTranslation service. Rather than measurin…