This dissertation compares the performance of Neural Machine Translation (NMT) and Artificial Intelligence (AI) tools in the context of Audiovisual Translation (AVT), specifically focusing on a Post-Editing (PE) analysis. The study is based on the documentary series "OJ25" produced by CourtTV, which covers the 1995 trial against O.J. Simpson. The project involved post-editing the NMT and AI-translated subtitles of selected interrogations and the full final episode, using OOONA for the spotting phase and Trados Studio for the PE process. The analysis is divided into qualitative and quantitative sections. The Translation Quality Assessment (TQA) employs the FAR methodology to categorize errors produced by the two tools. The quantitative analysis calculates the Translation Error Rate plus (TERp) to measure the post-editing effort for both NMT and AI. The findings highlight the distinct error patterns generated by the different systems, providing a measurable assessment of the required PE effort for each tool, contributing to the understanding of MT efficacy in specialized legal AVT.

Subtitling the Trial of the Century: The People v. O.J. Simpson

CROZZOLI, CHIARA
2024/2025

Abstract

This dissertation compares the performance of Neural Machine Translation (NMT) and Artificial Intelligence (AI) tools in the context of Audiovisual Translation (AVT), specifically focusing on a Post-Editing (PE) analysis. The study is based on the documentary series "OJ25" produced by CourtTV, which covers the 1995 trial against O.J. Simpson. The project involved post-editing the NMT and AI-translated subtitles of selected interrogations and the full final episode, using OOONA for the spotting phase and Trados Studio for the PE process. The analysis is divided into qualitative and quantitative sections. The Translation Quality Assessment (TQA) employs the FAR methodology to categorize errors produced by the two tools. The quantitative analysis calculates the Translation Error Rate plus (TERp) to measure the post-editing effort for both NMT and AI. The findings highlight the distinct error patterns generated by the different systems, providing a measurable assessment of the required PE effort for each tool, contributing to the understanding of MT efficacy in specialized legal AVT.
2024
Subtitling the Trial of the Century: The People v. O.J. Simpson
Subtitling
Legal English
AVT
Post-editing
File in questo prodotto:
File Dimensione Formato  
Crozzoli_Chiara.pdf

accesso aperto

Dimensione 8.86 MB
Formato Adobe PDF
8.86 MB Adobe PDF Visualizza/Apri

The text of this website © Università degli studi di Padova. Full Text are published under a non-exclusive license. Metadata are under a CC0 License

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/20.500.12608/100800