2018
DOI: 10.1007/978-3-319-75487-1_9
|View full text |Cite
|
Sign up to set email alerts
|

Genetic-Based Decoder for Statistical Machine Translation

Abstract: We propose a new algorithm for decoding on machine translation process. This approach is based on an evolutionary algorithm. We hope that this new method will constitute an alternative to Moses's decoder which is based on a beam search algorithm while the one we propose is based on the optimisation of a total solution. The results achieved are very encouraging in terms of measures and the proposed translations themselves are well built.

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1

Citation Types

0
0
0

Publication Types

Select...
1
1

Relationship

0
2

Authors

Journals

citations
Cited by 2 publications
(2 citation statements)
references
References 14 publications
0
0
0
Order By: Relevance
“…Since this approach is key to our work presented in part of Chapter 4 and in Chapter 5, we describe it with greater detail in Section 2.2.2. More recently, Douib et al (2016) use a genetic algorithm (Holland, 1973) for decoding full documents. This approach bears some resemblance to the one by Hardmeier et al (2012): the process starts by randomly generating a set of translations and then iteratively improving them by randomly exploring their neighborhoods.…”
Section: Document-level Statistical Machine Translationmentioning
confidence: 99%
See 1 more Smart Citation
“…Since this approach is key to our work presented in part of Chapter 4 and in Chapter 5, we describe it with greater detail in Section 2.2.2. More recently, Douib et al (2016) use a genetic algorithm (Holland, 1973) for decoding full documents. This approach bears some resemblance to the one by Hardmeier et al (2012): the process starts by randomly generating a set of translations and then iteratively improving them by randomly exploring their neighborhoods.…”
Section: Document-level Statistical Machine Translationmentioning
confidence: 99%
“…The ACO-based proposal we presented only obtains average results at sentence level, but further variants of the ACO metaheuristic are explored in the literature and their applicability to document-level MT might be more suitable and should be studied. Furthermore, other standard metaheuristics beyond the already tried simulated annealing (Hardmeier, 2014), genetic algorithms (Douib et al, 2016), and ACO could also be considered for guiding the document-aware decoding process.…”
Section: Future Workmentioning
confidence: 99%