Introduction
Propio created and trained domain-specific Neural Machine Translation (NMT) engines and tested their effectiveness across multiple domains. The results? A methodology that can be applied to industries and clients for an improved quality output – even with non-traditional languages.
Propio provides Neural Machine Translation (NMT) in over 130 translation directions for global clients across multiple markets and domains.
Each engine is designed to provide translations within the scope of a generic subject area. In order to integrate NMT in the production workflow, domain engines needed to be created, so Propio built domain-specific engines pertaining to medical devices, pharmaceutical, heavy machinery, IFU-DFU and Healthcare.
The Challenge
The client partnered with Propio to determine what methods of translation would best fit each content need and ensure accuracy of the output. Propio created a customized technology and translation plan to achieve the following:
- Maintaining compliance with industry regulations.
- Streamlining internal communication among its global workforce
- Reducing costs for non-critical content
- Decreasing time to market
Propio determined the utilization of Neural Machine Translation (NMT) would help the client achieve the increased speed it required.
To maintain the quality, Propio was able to build a client-specific mix of human and machine translation that would ensure consistency and accuracy across languages.
For the content that needed to meet the steep regulatory and quality requirements, Propio implemented a combined machine translation, human translation and human editing solution. This unique blend effectively reduced costs by using a seamless workflow integrated with a customized NMT solution that opened the door to greater efficiencies.
The Solution
Propio created custom domain-specific NMT engines that increased the quality scores of the translation output.
Translation quality is scored using multiple evaluation methods, including the standard BLEU quality score which is a computer-generated metric that analyzes the results by comparing the NMT file to a reference file. Propio also uses TER (automated edit distance scoring) and independent Distance Scoring performed by a team of linguists to determine and validate quality scoring of NMT.
When creating the new engines, Propio’s NMT team followed the standard engine customization process and obtained significant BLEU score increases compared to the score of the in-market engine and the scores of the new domain engines.
Proven Results
By utilizing a process and technology mix that combined the quality benefits of human translation with the expedience of NMT, the client experienced dramatic results. For its regulated content, the company saved 25 percent on translation.
In addition, its internal communications costs decreased by approximately 44 percent with turnaround time reduced by 45 percent.
