# Adaptation of Machine Translation Models with Back-translated Data using Transductive Data Selection Methods

## TL;DR

This paper explores the use of transductive data selection methods to adapt neural machine translation models with back-translated synthetic data, demonstrating that selective adaptation improves translation quality despite noise.

## Contribution

It introduces the application of INR and FDA data selection methods to synthetic back-translated data for NMT model adaptation, addressing challenges of noise and test set relevance.

## Key findings

- Data selection improves NMT adaptation with synthetic data.
- Transductive methods effectively identify relevant synthetic sentences.
- Synthetic data adaptation yields better translation performance.

## Abstract

Data selection has proven its merit for improving Neural Machine Translation (NMT), when applied to authentic data. But the benefit of using synthetic data in NMT training, produced by the popular back-translation technique, raises the question if data selection could also be useful for synthetic data? In this work we use Infrequent N-gram Recovery (INR) and Feature Decay Algorithms (FDA), two transductive data selection methods to obtain subsets of sentences from synthetic data. These methods ensure that selected sentences share n-grams with the test set so the NMT model can be adapted to translate it. Performing data selection on back-translated data creates new challenges as the source-side may contain noise originated by the model used in the back-translation. Hence, finding n-grams present in the test set become more difficult. Despite that, in our work we show that adapting a model with a selection of synthetic data is an useful approach.