Matches in SemOpenAlex for { <https://semopenalex.org/work/W2032942114> ?p ?o ?g. }
- W2032942114 endingPage "29" @default.
- W2032942114 startingPage "1" @default.
- W2032942114 abstract "We explore the use of morph-based language models in large-vocabulary continuous-speech recognition systems across four so-called morphologically rich languages: Finnish, Estonian, Turkish, and Egyptian Colloquial Arabic. The morphs are subword units discovered in an unsupervised, data-driven way using the Morfessor algorithm. By estimating n -gram language models over sequences of morphs instead of words, the quality of the language model is improved through better vocabulary coverage and reduced data sparsity. Standard word models suffer from high out-of-vocabulary (OOV) rates, whereas the morph models can recognize previously unseen word forms by concatenating morphs. It is shown that the morph models do perform fairly well on OOVs without compromising the recognition accuracy on in-vocabulary words. The Arabic experiment constitutes the only exception since here the standard word model outperforms the morph model. Differences in the datasets and the amount of data are discussed as a plausible explanation." @default.
- W2032942114 created "2016-06-24" @default.
- W2032942114 creator A5030484123 @default.
- W2032942114 creator A5032284977 @default.
- W2032942114 creator A5040694307 @default.
- W2032942114 creator A5043424064 @default.
- W2032942114 creator A5046171140 @default.
- W2032942114 creator A5055086464 @default.
- W2032942114 creator A5060979948 @default.
- W2032942114 creator A5061159001 @default.
- W2032942114 creator A5085765257 @default.
- W2032942114 creator A5089859626 @default.
- W2032942114 date "2007-12-01" @default.
- W2032942114 modified "2023-09-26" @default.
- W2032942114 title "Morph-based speech recognition and modeling of out-of-vocabulary words across languages" @default.
- W2032942114 cites W1983311927 @default.
- W2032942114 cites W2010910318 @default.
- W2032942114 cites W2042783153 @default.
- W2032942114 cites W2050938027 @default.
- W2032942114 cites W2053306448 @default.
- W2032942114 cites W2056250865 @default.
- W2032942114 cites W2069712814 @default.
- W2032942114 cites W2074546930 @default.
- W2032942114 cites W2101711363 @default.
- W2032942114 cites W2103589071 @default.
- W2032942114 cites W2117621558 @default.
- W2032942114 cites W2122228338 @default.
- W2032942114 cites W2141684702 @default.
- W2032942114 cites W2150144720 @default.
- W2032942114 cites W2158195707 @default.
- W2032942114 cites W4251556668 @default.
- W2032942114 doi "https://doi.org/10.1145/1322391.1322394" @default.
- W2032942114 hasPublicationYear "2007" @default.
- W2032942114 type Work @default.
- W2032942114 sameAs 2032942114 @default.
- W2032942114 citedByCount "105" @default.
- W2032942114 countsByYear W20329421142012 @default.
- W2032942114 countsByYear W20329421142013 @default.
- W2032942114 countsByYear W20329421142014 @default.
- W2032942114 countsByYear W20329421142015 @default.
- W2032942114 countsByYear W20329421142016 @default.
- W2032942114 countsByYear W20329421142017 @default.
- W2032942114 countsByYear W20329421142018 @default.
- W2032942114 countsByYear W20329421142019 @default.
- W2032942114 countsByYear W20329421142020 @default.
- W2032942114 countsByYear W20329421142021 @default.
- W2032942114 countsByYear W20329421142022 @default.
- W2032942114 countsByYear W20329421142023 @default.
- W2032942114 crossrefType "journal-article" @default.
- W2032942114 hasAuthorship W2032942114A5030484123 @default.
- W2032942114 hasAuthorship W2032942114A5032284977 @default.
- W2032942114 hasAuthorship W2032942114A5040694307 @default.
- W2032942114 hasAuthorship W2032942114A5043424064 @default.
- W2032942114 hasAuthorship W2032942114A5046171140 @default.
- W2032942114 hasAuthorship W2032942114A5055086464 @default.
- W2032942114 hasAuthorship W2032942114A5060979948 @default.
- W2032942114 hasAuthorship W2032942114A5061159001 @default.
- W2032942114 hasAuthorship W2032942114A5085765257 @default.
- W2032942114 hasAuthorship W2032942114A5089859626 @default.
- W2032942114 hasConcept C117884012 @default.
- W2032942114 hasConcept C137293760 @default.
- W2032942114 hasConcept C138885662 @default.
- W2032942114 hasConcept C154945302 @default.
- W2032942114 hasConcept C204321447 @default.
- W2032942114 hasConcept C2777601683 @default.
- W2032942114 hasConcept C2778243841 @default.
- W2032942114 hasConcept C2781121862 @default.
- W2032942114 hasConcept C28490314 @default.
- W2032942114 hasConcept C41008148 @default.
- W2032942114 hasConcept C41895202 @default.
- W2032942114 hasConcept C90805587 @default.
- W2032942114 hasConcept C96455323 @default.
- W2032942114 hasConceptScore W2032942114C117884012 @default.
- W2032942114 hasConceptScore W2032942114C137293760 @default.
- W2032942114 hasConceptScore W2032942114C138885662 @default.
- W2032942114 hasConceptScore W2032942114C154945302 @default.
- W2032942114 hasConceptScore W2032942114C204321447 @default.
- W2032942114 hasConceptScore W2032942114C2777601683 @default.
- W2032942114 hasConceptScore W2032942114C2778243841 @default.
- W2032942114 hasConceptScore W2032942114C2781121862 @default.
- W2032942114 hasConceptScore W2032942114C28490314 @default.
- W2032942114 hasConceptScore W2032942114C41008148 @default.
- W2032942114 hasConceptScore W2032942114C41895202 @default.
- W2032942114 hasConceptScore W2032942114C90805587 @default.
- W2032942114 hasConceptScore W2032942114C96455323 @default.
- W2032942114 hasFunder F4320332180 @default.
- W2032942114 hasIssue "1" @default.
- W2032942114 hasLocation W20329421141 @default.
- W2032942114 hasOpenAccess W2032942114 @default.
- W2032942114 hasPrimaryLocation W20329421141 @default.
- W2032942114 hasRelatedWork W130046785 @default.
- W2032942114 hasRelatedWork W1963734729 @default.
- W2032942114 hasRelatedWork W2008468404 @default.
- W2032942114 hasRelatedWork W2043693083 @default.
- W2032942114 hasRelatedWork W2100172690 @default.
- W2032942114 hasRelatedWork W2137397532 @default.
- W2032942114 hasRelatedWork W3196833733 @default.
- W2032942114 hasRelatedWork W4301311969 @default.