Compositional Morphology for Word Representations and Language Modelling
This paper presents a scalable method for integrating compositional morphological representations into a vector-based probabilistic language model. Our approach is evaluated in the context of log-bilinear language models, rendered suitably efficient for implementation inside a machine translation de...
| Hoofdauteurs: | , |
|---|---|
| Formaat: | Conference item |
| Gepubliceerd in: |
Beijing‚ China
2014
|
| _version_ | 1826257535943835648 |
|---|---|
| author | Botha, J Blunsom, P |
| author_facet | Botha, J Blunsom, P |
| author_sort | Botha, J |
| collection | OXFORD |
| description | This paper presents a scalable method for integrating compositional morphological representations into a vector-based probabilistic language model. Our approach is evaluated in the context of log-bilinear language models, rendered suitably efficient for implementation inside a machine translation decoder by factoring the vocabulary. We perform both intrinsic and extrinsic evaluations, presenting results on a range of languages which demonstrate that our model learns morphological representations that both perform well on word similarity tasks and lead to substantial reductions in perplexity. When used for translation into morphologically rich languages with large vocabularies, our models obtain improvements of up to 1.2 BLEU points relative to a baseline system using back-off n-gram models. |
| first_indexed | 2024-03-06T18:19:46Z |
| format | Conference item |
| id | oxford-uuid:05dccb3e-af94-4b4c-b112-fce79cacd12b |
| institution | University of Oxford |
| last_indexed | 2024-03-06T18:19:46Z |
| publishDate | 2014 |
| publisher | Beijing‚ China |
| record_format | dspace |
| spelling | oxford-uuid:05dccb3e-af94-4b4c-b112-fce79cacd12b2022-03-26T08:59:32ZCompositional Morphology for Word Representations and Language ModellingConference itemhttp://purl.org/coar/resource_type/c_5794uuid:05dccb3e-af94-4b4c-b112-fce79cacd12bDepartment of Computer ScienceBeijing‚ China2014Botha, JBlunsom, PThis paper presents a scalable method for integrating compositional morphological representations into a vector-based probabilistic language model. Our approach is evaluated in the context of log-bilinear language models, rendered suitably efficient for implementation inside a machine translation decoder by factoring the vocabulary. We perform both intrinsic and extrinsic evaluations, presenting results on a range of languages which demonstrate that our model learns morphological representations that both perform well on word similarity tasks and lead to substantial reductions in perplexity. When used for translation into morphologically rich languages with large vocabularies, our models obtain improvements of up to 1.2 BLEU points relative to a baseline system using back-off n-gram models. |
| spellingShingle | Botha, J Blunsom, P Compositional Morphology for Word Representations and Language Modelling |
| title | Compositional Morphology for Word Representations and Language Modelling |
| title_full | Compositional Morphology for Word Representations and Language Modelling |
| title_fullStr | Compositional Morphology for Word Representations and Language Modelling |
| title_full_unstemmed | Compositional Morphology for Word Representations and Language Modelling |
| title_short | Compositional Morphology for Word Representations and Language Modelling |
| title_sort | compositional morphology for word representations and language modelling |
| work_keys_str_mv | AT bothaj compositionalmorphologyforwordrepresentationsandlanguagemodelling AT blunsomp compositionalmorphologyforwordrepresentationsandlanguagemodelling |