Results 31 to 40 of about 133,026 (278)
, 2023 Current research in zero-shot translation is plagued by several issues such as high compute requirements, increased training time and off target translations. Proposed remedies often come at the cost of additional data or compute requirements. Pivot based neural machine translation is preferred over a single-encoder model for most settings despite the ...
Ebadulla, Danish +4 more
openaire +2 more sources
On the reference of zero subjects
Lietuvių Kalba, 2014 The article deals with the issue of the reference of zero subjects which has been very scarcely examined in Lithuanian linguistics. The first part of the article presents a review of the theoretical tools to analyse zero elements and its key theoretical ...
Simona Akamauskaitė, Vytautas Kardelis
doaj +1 more source
Constituent Parsing as Sequence Labeling [PDF]
, 2018 We introduce a method to reduce constituent parsing to sequence labeling. For each word w_t, it generates a label that encodes: (1) the number of ancestors in the tree that the words w_t and w_{t+1} have in common, and (2) the nonterminal symbol at the ...
Gómez-Rodríguez, Carlos +1 more
core +2 more sources
Nominalization and alternations in biomedical language. [PDF]
PLoS ONE, 2008 This paper presents data on alternations in the argument structure of common domain-specific verbs and their associated verbal nominalizations in the PennBioIE corpus.
K Bretonnel Cohen +2 more
doaj +1 more source
ENTROPY OF LANGUAGE SYSTEM AS MAIN DEVELOPMENT INDICATOR
Sovremennye Issledovaniâ Socialʹnyh Problem, 2014 Article is devoted to the research of language system entropy. It is one of the main concept of synergetics and synergy. Entropy can be applied to description of language processes and detection of functioning and development features of language ...
Ирина Михайловна Некипелова
doaj +1 more source
Zero-shot Generative Linguistic Steganography
Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers)Generative linguistic steganography attempts to hide secret messages into covertext. Previous studies have generally focused on the statistical differences between the covertext and stegotext, however, ill-formed stegotext can readily be identified by humans.
Lin, Ke +3 more
openaire +2 more sources
Crossings as a side effect of dependency lengths [PDF]
, 2016 The syntactic structure of sentences exhibits a striking regularity: dependencies tend to not cross when drawn above the sentence. We investigate two competing explanations.
Bick +30 more
core +4 more sources
Multilingual NMT with a language-independent attention bridge
, 2018 In this paper, we propose a multilingual encoder-decoder architecture capable of obtaining multilingual sentence representations by means of incorporating an intermediate {\em attention bridge} that is shared across all languages.
Creutz, Mathias +3 more
core +1 more source
British Educational Research Journal, EarlyView., 2023 Abstract Children's relationship with time in preschools is an under‐researched area. Young children rarely know how to measure time using a clock, but their experiences of time may contribute to understanding children's well‐being and debates about quality in preschools.
Kristín Dýrfjörð +3 more
wiley +1 more source
PHRASEMES IN ANDRIĆ’S SHORT STORIES AND THEIR TRANSLATION EQUIVALENTS IN GERMAN LANGUAGE [PDF]
Hum, 2012 Phraseology as a linguistic discipline does not have long tradition, but that does not alleviate its importance either within one language or in contrasting two languages.
Magdalena Ramljak, Marija Musa
doaj