A Reinforcement Learning Approach to Interactive-Predictive Neural Machine Translation

05/03/2018
by   Tsz Kin Lam, et al.
0

We present an approach to interactive-predictive neural machine translation that attempts to reduce human effort from three directions: Firstly, instead of requiring humans to select, correct, or delete segments, we employ the idea of learning from human reinforcements in form of judgments on the quality of partial translations. Secondly, human effort is further reduced by using the entropy of word predictions as uncertainty criterion to trigger feedback requests. Lastly, online updates of the model parameters after every interaction allow the model to adapt quickly. We show in simulation experiments that reward signals on partial translations significantly improve character F-score and BLEU compared to feedback on full translations only, while human effort can be reduced to an average number of 5 feedback requests for every input.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
07/04/2019

Interactive-Predictive Neural Machine Translation through Reinforcement and Imitation

We propose an interactive-predictive neural machine translation framewor...
research
01/18/2016

Bandit Structured Prediction for Learning from Partial Feedback in Statistical Machine Translation

We present an approach to structured prediction from bandit feedback, ca...
research
02/10/2018

Online Learning for Effort Reduction in Interactive Neural Machine Translation

Neural machine translation systems require large amounts of training dat...
research
04/23/2020

Correct Me If You Can: Learning from Error Corrections and Markings

Sequence-to-sequence learning involves a trade-off between signal streng...
research
07/24/2017

Reinforcement Learning for Bandit Neural Machine Translation with Simulated Human Feedback

Machine translation is a natural candidate problem for reinforcement lea...
research
03/28/2019

Train, Sort, Explain: Learning to Diagnose Translation Models

Evaluating translation models is a trade-off between effort and detail. ...
research
05/20/2019

A Neural, Interactive-predictive System for Multimodal Sequence to Sequence Tasks

We present a demonstration of a neural interactive-predictive system for...

Please sign up or login with your details

Forgot password? Click here to reset