Matches in SemOpenAlex for { <https://semopenalex.org/work/W2980470892> ?p ?o ?g. }
Showing items 1 to 94 of
94
with 100 items per page.
- W2980470892 abstract "Weighted finite automata (WFA) are often used to represent probabilistic models, such as n-gram language models, since they are efficient for recognition tasks in time and space. The probabilistic source to be represented as a WFA, however, may come in many forms. Given a generic probabilistic model over sequences, we propose an algorithm to approximate it as a weighted finite automaton such that the Kullback-Leibler divergence between the source model and the WFA target model is minimized. The proposed algorithm involves a counting step and a difference of convex optimization, both of which can be performed efficiently. We demonstrate the usefulness of our approach on some tasks including distilling n-gram models from neural models." @default.
- W2980470892 created "2019-10-25" @default.
- W2980470892 creator A5014137885 @default.
- W2980470892 creator A5020068498 @default.
- W2980470892 creator A5080091001 @default.
- W2980470892 creator A5085885345 @default.
- W2980470892 date "2019-01-01" @default.
- W2980470892 modified "2023-09-25" @default.
- W2980470892 title "Distilling weighted finite automata from arbitrary probabilistic models" @default.
- W2980470892 cites W138349765 @default.
- W2980470892 cites W142858687 @default.
- W2980470892 cites W1486103693 @default.
- W2980470892 cites W1585876329 @default.
- W2980470892 cites W1629199664 @default.
- W2980470892 cites W164910094 @default.
- W2980470892 cites W1797288984 @default.
- W2980470892 cites W1818782312 @default.
- W2980470892 cites W1895481600 @default.
- W2980470892 cites W1988857989 @default.
- W2980470892 cites W1993557076 @default.
- W2980470892 cites W2009570821 @default.
- W2980470892 cites W2049633694 @default.
- W2980470892 cites W2082863623 @default.
- W2980470892 cites W2099964107 @default.
- W2980470892 cites W2103085228 @default.
- W2980470892 cites W2110415041 @default.
- W2980470892 cites W2118990292 @default.
- W2980470892 cites W2125528794 @default.
- W2980470892 cites W2134237567 @default.
- W2980470892 cites W2137853299 @default.
- W2980470892 cites W2140756703 @default.
- W2980470892 cites W2154932389 @default.
- W2980470892 cites W2158195707 @default.
- W2980470892 cites W2163489213 @default.
- W2980470892 cites W2171842071 @default.
- W2980470892 cites W2293770964 @default.
- W2980470892 cites W2295078202 @default.
- W2980470892 cites W2402268235 @default.
- W2980470892 cites W2535838896 @default.
- W2980470892 cites W2605710880 @default.
- W2980470892 cites W2746762542 @default.
- W2980470892 cites W2810821719 @default.
- W2980470892 cites W2900120080 @default.
- W2980470892 cites W62305045 @default.
- W2980470892 cites W81972919 @default.
- W2980470892 cites W2037942319 @default.
- W2980470892 doi "https://doi.org/10.18653/v1/w19-3112" @default.
- W2980470892 hasPublicationYear "2019" @default.
- W2980470892 type Work @default.
- W2980470892 sameAs 2980470892 @default.
- W2980470892 citedByCount "5" @default.
- W2980470892 countsByYear W29804708922019 @default.
- W2980470892 countsByYear W29804708922022 @default.
- W2980470892 crossrefType "proceedings-article" @default.
- W2980470892 hasAuthorship W2980470892A5014137885 @default.
- W2980470892 hasAuthorship W2980470892A5020068498 @default.
- W2980470892 hasAuthorship W2980470892A5080091001 @default.
- W2980470892 hasAuthorship W2980470892A5085885345 @default.
- W2980470892 hasBestOaLocation W29804708921 @default.
- W2980470892 hasConcept C112505250 @default.
- W2980470892 hasConcept C11413529 @default.
- W2980470892 hasConcept C116248031 @default.
- W2980470892 hasConcept C154945302 @default.
- W2980470892 hasConcept C167822520 @default.
- W2980470892 hasConcept C174327141 @default.
- W2980470892 hasConcept C41008148 @default.
- W2980470892 hasConcept C49937458 @default.
- W2980470892 hasConcept C80444323 @default.
- W2980470892 hasConceptScore W2980470892C112505250 @default.
- W2980470892 hasConceptScore W2980470892C11413529 @default.
- W2980470892 hasConceptScore W2980470892C116248031 @default.
- W2980470892 hasConceptScore W2980470892C154945302 @default.
- W2980470892 hasConceptScore W2980470892C167822520 @default.
- W2980470892 hasConceptScore W2980470892C174327141 @default.
- W2980470892 hasConceptScore W2980470892C41008148 @default.
- W2980470892 hasConceptScore W2980470892C49937458 @default.
- W2980470892 hasConceptScore W2980470892C80444323 @default.
- W2980470892 hasLocation W29804708921 @default.
- W2980470892 hasOpenAccess W2980470892 @default.
- W2980470892 hasPrimaryLocation W29804708921 @default.
- W2980470892 hasRelatedWork W1587826772 @default.
- W2980470892 hasRelatedWork W1762969849 @default.
- W2980470892 hasRelatedWork W1979390054 @default.
- W2980470892 hasRelatedWork W1988501450 @default.
- W2980470892 hasRelatedWork W2044753318 @default.
- W2980470892 hasRelatedWork W2076986528 @default.
- W2980470892 hasRelatedWork W2101451117 @default.
- W2980470892 hasRelatedWork W2142578041 @default.
- W2980470892 hasRelatedWork W2159449605 @default.
- W2980470892 hasRelatedWork W3095160120 @default.
- W2980470892 isParatext "false" @default.
- W2980470892 isRetracted "false" @default.
- W2980470892 magId "2980470892" @default.
- W2980470892 workType "article" @default.