Matches in SemOpenAlex for { <https://semopenalex.org/work/W4320341756> ?p ?o ?g. }
Showing items 1 to 83 of
83
with 100 items per page.
- W4320341756 abstract "The distributed representation of symbols is one of the key technologies in machine learning systems today, playing a pivotal role in modern natural language processing. Traditional word embeddings associate a separate vector with each word. While this approach is simple and leads to good performance, it requires a lot of memory for representing a large vocabulary. To reduce the memory footprint, the default embedding layer in spaCy is a hash embeddings layer. It is a stochastic approximation of traditional embeddings that provides unique vectors for a large number of words without explicitly storing a separate vector for each of them. To be able to compute meaningful representations for both known and unknown words, hash embeddings represent each word as a summary of the normalized word form, subword information and word shape. Together, these features produce a multi-embedding of a word. In this technical report we lay out a bit of history and introduce the embedding methods in spaCy in detail. Second, we critically evaluate the hash embedding architecture with multi-embeddings on Named Entity Recognition datasets from a variety of domains and languages. The experiments validate most key design choices behind spaCy's embedders, but we also uncover a few surprising results." @default.
- W4320341756 created "2023-02-13" @default.
- W4320341756 creator A5018138946 @default.
- W4320341756 creator A5029159426 @default.
- W4320341756 creator A5046446401 @default.
- W4320341756 creator A5047556774 @default.
- W4320341756 creator A5081770499 @default.
- W4320341756 creator A5084996762 @default.
- W4320341756 date "2022-12-19" @default.
- W4320341756 modified "2023-09-23" @default.
- W4320341756 title "Multi hash embeddings in spaCy" @default.
- W4320341756 doi "https://doi.org/10.48550/arxiv.2212.09255" @default.
- W4320341756 hasPublicationYear "2022" @default.
- W4320341756 type Work @default.
- W4320341756 citedByCount "0" @default.
- W4320341756 crossrefType "posted-content" @default.
- W4320341756 hasAuthorship W4320341756A5018138946 @default.
- W4320341756 hasAuthorship W4320341756A5029159426 @default.
- W4320341756 hasAuthorship W4320341756A5046446401 @default.
- W4320341756 hasAuthorship W4320341756A5047556774 @default.
- W4320341756 hasAuthorship W4320341756A5081770499 @default.
- W4320341756 hasAuthorship W4320341756A5084996762 @default.
- W4320341756 hasBestOaLocation W43203417561 @default.
- W4320341756 hasConcept C111472728 @default.
- W4320341756 hasConcept C136197465 @default.
- W4320341756 hasConcept C138885662 @default.
- W4320341756 hasConcept C154945302 @default.
- W4320341756 hasConcept C17744445 @default.
- W4320341756 hasConcept C199539241 @default.
- W4320341756 hasConcept C204321447 @default.
- W4320341756 hasConcept C2524010 @default.
- W4320341756 hasConcept C26517878 @default.
- W4320341756 hasConcept C2776359362 @default.
- W4320341756 hasConcept C2777462759 @default.
- W4320341756 hasConcept C2777601683 @default.
- W4320341756 hasConcept C2780586882 @default.
- W4320341756 hasConcept C33923547 @default.
- W4320341756 hasConcept C38652104 @default.
- W4320341756 hasConcept C41008148 @default.
- W4320341756 hasConcept C41608201 @default.
- W4320341756 hasConcept C41895202 @default.
- W4320341756 hasConcept C80444323 @default.
- W4320341756 hasConcept C90805587 @default.
- W4320341756 hasConcept C94625758 @default.
- W4320341756 hasConcept C99138194 @default.
- W4320341756 hasConceptScore W4320341756C111472728 @default.
- W4320341756 hasConceptScore W4320341756C136197465 @default.
- W4320341756 hasConceptScore W4320341756C138885662 @default.
- W4320341756 hasConceptScore W4320341756C154945302 @default.
- W4320341756 hasConceptScore W4320341756C17744445 @default.
- W4320341756 hasConceptScore W4320341756C199539241 @default.
- W4320341756 hasConceptScore W4320341756C204321447 @default.
- W4320341756 hasConceptScore W4320341756C2524010 @default.
- W4320341756 hasConceptScore W4320341756C26517878 @default.
- W4320341756 hasConceptScore W4320341756C2776359362 @default.
- W4320341756 hasConceptScore W4320341756C2777462759 @default.
- W4320341756 hasConceptScore W4320341756C2777601683 @default.
- W4320341756 hasConceptScore W4320341756C2780586882 @default.
- W4320341756 hasConceptScore W4320341756C33923547 @default.
- W4320341756 hasConceptScore W4320341756C38652104 @default.
- W4320341756 hasConceptScore W4320341756C41008148 @default.
- W4320341756 hasConceptScore W4320341756C41608201 @default.
- W4320341756 hasConceptScore W4320341756C41895202 @default.
- W4320341756 hasConceptScore W4320341756C80444323 @default.
- W4320341756 hasConceptScore W4320341756C90805587 @default.
- W4320341756 hasConceptScore W4320341756C94625758 @default.
- W4320341756 hasConceptScore W4320341756C99138194 @default.
- W4320341756 hasLocation W43203417561 @default.
- W4320341756 hasOpenAccess W4320341756 @default.
- W4320341756 hasPrimaryLocation W43203417561 @default.
- W4320341756 hasRelatedWork W2374192425 @default.
- W4320341756 hasRelatedWork W2390245496 @default.
- W4320341756 hasRelatedWork W2573169100 @default.
- W4320341756 hasRelatedWork W2949267551 @default.
- W4320341756 hasRelatedWork W3086845375 @default.
- W4320341756 hasRelatedWork W3095760691 @default.
- W4320341756 hasRelatedWork W3107474891 @default.
- W4320341756 hasRelatedWork W3146034405 @default.
- W4320341756 hasRelatedWork W3158961393 @default.
- W4320341756 hasRelatedWork W4287629333 @default.
- W4320341756 isParatext "false" @default.
- W4320341756 isRetracted "false" @default.
- W4320341756 workType "article" @default.