Confidence through Attention

10/10/2017
by   Matīss Rikters, et al.
0

Attention distributions of the generated translations are a useful bi-product of attention-based recurrent neural network translation models and can be treated as soft alignments between the input and output tokens. In this work, we use attention distributions as a confidence metric for output translations. We present two strategies of using the attention distributions: filtering out bad translations from a large back-translated corpus, and selecting the best translation in a hybrid setup of two different translation systems. While manual evaluation indicated only a weak correlation between our confidence score and human judgments, the use-cases showed improvements of up to 2.22 BLEU points for filtering and 0.99 points for hybrid translation, tested on English<->German and English<->Latvian translation.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
08/08/2018

Debugging Neural Machine Translations

In this paper, we describe a tool for debugging the output and attention...
research
06/09/2016

Edinburgh Neural Machine Translation Systems for WMT 16

We participated in the WMT 2016 shared news translation task by building...
research
07/15/2019

Facebook FAIR's WMT19 News Translation Task Submission

This paper describes Facebook FAIR's submission to the WMT19 shared news...
research
12/19/2016

Neural Machine Translation from Simplified Translations

Text simplification aims at reducing the lexical, grammatical and struct...
research
10/20/2016

Iterative Refinement for Machine Translation

Existing machine translation decoding algorithms generate translations i...
research
01/05/2015

Chasing the Ghosts of Ibsen: A computational stylistic analysis of drama in translation

Research into the stylistic properties of translations is an issue which...
research
07/28/2023

Multilingual Tourist Assistance using ChatGPT: Comparing Capabilities in Hindi, Telugu, and Kannada

This research investigates the effectiveness of ChatGPT, an AI language ...

Please sign up or login with your details

Forgot password? Click here to reset