Nearest Neighbor Machine Translation
This approach addresses the challenge of enhancing translation quality and adaptability for users in multilingual and domain-specific contexts, offering a novel method that is incremental in combining existing techniques.
The paper tackles the problem of improving machine translation performance by introducing k-nearest-neighbor machine translation (kNN-MT), which uses a nearest neighbor classifier over cached examples without additional training, resulting in consistent improvements such as a 1.5 BLEU gain for German-English translation and up to 9.2 BLEU for domain adaptation.
We introduce $k$-nearest-neighbor machine translation ($k$NN-MT), which predicts tokens with a nearest neighbor classifier over a large datastore of cached examples, using representations from a neural translation model for similarity search. This approach requires no additional training and scales to give the decoder direct access to billions of examples at test time, resulting in a highly expressive model that consistently improves performance across many settings. Simply adding nearest neighbor search improves a state-of-the-art German-English translation model by 1.5 BLEU. $k$NN-MT allows a single model to be adapted to diverse domains by using a domain-specific datastore, improving results by an average of 9.2 BLEU over zero-shot transfer, and achieving new state-of-the-art results -- without training on these domains. A massively multilingual model can also be specialized for particular language pairs, with improvements of 3 BLEU for translating from English into German and Chinese. Qualitatively, $k$NN-MT is easily interpretable; it combines source and target context to retrieve highly relevant examples.