id	author	title	date	pages	extension	mime	words	sentences	flesch	summary	cache	txt
work_rxaui7o4kzczlmk53n3amxhe54	Benjamin Marie	Phrase Table Induction Using In-Domain Monolingual Data for Domain Adaptation in Statistical Machine Translation	2017	14	.pdf	application/pdf	9290	706	65	We present a new framework to induce an indomain phrase table from in-domain monolingual data that can be used to adapt a generaldomain statistical machine translation system We also conducted an error analysis that showed the induced phrase tables proposed useful translations, especially for words and phrases unseen data or highly comparable corpora, our method induces an in-domain phrase table from unaligned indomain monolingual data through a three-step pro• dealing with source and target phrases of arbitrary length collected from in-domain monolingual data, (2014) uses only phrases from the target side of their parallel data and their morphological variants ranked and pruned according to the forward lexical translation probabilities given by their in the general-domain parallel data, lexical translation probabilities may be useful to score candidate This section demonstrates the impact of the induced phrase tables in translating in-domain texts For each pair of domain and translation direction, sets of source and target phrases were extracted	./cache/work_rxaui7o4kzczlmk53n3amxhe54.pdf	./txt/work_rxaui7o4kzczlmk53n3amxhe54.txt