TY - JOUR
TI - A Translation Evaluation Function based on Neural Network
AU - Douib, Ameur
AU - Langlois, David
AU - Smaïli, Kamel
TI - A Translation Evaluation Function based on Neural Network
AB - 
	In this paper, we study the feasibility of using a neural network to learn a fitness function for a machine translation system based on a genetic algorithm termed GAMaT. The neural network is learned on  features extracted from pairs of source sentences and their translations. The fitness function is trained in order to estimate the BLEU of a translation as precisely as possible. The estimator has been trained on a corpus of more than 1.3 million data. The performance is very promising: the difference between the real BLEU and the one given by the estimator is equal to 0.12 in terms of Mean Absolute Error.
VL - 2016
IS - Volume 25
PY - 2017
SN - 1732-3916
C1 - 2083-8476
SP - 139
EP - 151
DO - 10.4467/20838476SI.16.011.6192
UR - https://ejournals.eu/en/journal/schedae-informaticae/article/a-translation-evaluation-function-based-on-neural-network
KW - Statistical Machine Translation
KW - Genetic algorithm
KW - Quality estimation
KW - Neural network