Journal article Open Access
MinhQuang Pham; Josep Crego; François Yvon
{ "files": [ { "links": { "self": "https://zenodo.org/api/files/bdb72c2b-a896-4785-8e62-955eb94ce9e6/main-2327-PhamMinhQuang.pdf" }, "checksum": "md5:c2db23c132297b90a385481171fff2c2", "bucket": "bdb72c2b-a896-4785-8e62-955eb94ce9e6", "key": "main-2327-PhamMinhQuang.pdf", "type": "pdf", "size": 284742 } ], "owners": [ 196497 ], "doi": "10.5281/zenodo.4537184", "stats": { "version_unique_downloads": 23.0, "unique_views": 29.0, "views": 42.0, "version_views": 42.0, "unique_downloads": 23.0, "version_unique_views": 29.0, "volume": 7118550.0, "version_downloads": 25.0, "downloads": 25.0, "version_volume": 7118550.0 }, "links": { "doi": "https://doi.org/10.5281/zenodo.4537184", "conceptdoi": "https://doi.org/10.5281/zenodo.4537183", "bucket": "https://zenodo.org/api/files/bdb72c2b-a896-4785-8e62-955eb94ce9e6", "conceptbadge": "https://zenodo.org/badge/doi/10.5281/zenodo.4537183.svg", "html": "https://zenodo.org/record/4537184", "latest_html": "https://zenodo.org/record/4537184", "badge": "https://zenodo.org/badge/doi/10.5281/zenodo.4537184.svg", "latest": "https://zenodo.org/api/records/4537184" }, "conceptdoi": "10.5281/zenodo.4537183", "created": "2021-02-12T10:39:44.763783+00:00", "updated": "2021-02-14T08:12:11.832829+00:00", "conceptrecid": "4537183", "revision": 4, "id": 4537184, "metadata": { "access_right_category": "success", "doi": "10.5281/zenodo.4537184", "description": "<p>When building machine translation systems, one often needs to make the best out of heterogeneous sets of parallel data in training, and to robustly handle inputs from un-expected domains in testing. This multi-domain scenario has attracted a lot of recent work, that fall under the general umbrella of transfer learning. In this study, we revisit multi-domain machine translation, with the aim to formulate the motivations for developing such systems and the associated expectations with respect to performance. Our experiments with a large sample of multi-domain systems show that most of these expectations are hardly met and suggest that further work is needed to better analyze the current behaviour of multi-domain systems and to make them fully hold their promises.</p>", "language": "eng", "title": "Revisiting Multi-Domain Machine Translation", "license": { "id": "CC-BY-4.0" }, "journal": { "volume": "9", "title": "Transactions of the Association for Computational Linguistics" }, "relations": { "version": [ { "count": 1, "index": 0, "parent": { "pid_type": "recid", "pid_value": "4537183" }, "is_last": true, "last_child": { "pid_type": "recid", "pid_value": "4537184" } } ] }, "communities": [ { "id": "787061" } ], "grants": [ { "code": "787061", "links": { "self": "https://zenodo.org/api/grants/10.13039/501100000780::787061" }, "title": "Advanced tools for fighting oNline Illegal TrAfficking", "acronym": "ANITA", "program": "H2020", "funder": { "doi": "10.13039/501100000780", "acronyms": [], "name": "European Commission", "links": { "self": "https://zenodo.org/api/funders/10.13039/501100000780" } } } ], "keywords": [ "Neural Machine Translation", "Multi-domain MT", "Domain Adaptation" ], "publication_date": "2021-02-12", "creators": [ { "affiliation": "SYSTRAN, LIMSI/CNRS", "name": "MinhQuang Pham" }, { "affiliation": "SYSTRAN", "name": "Josep Crego" }, { "affiliation": "LIMSI/CNRS", "name": "Fran\u00e7ois Yvon" } ], "access_right": "open", "resource_type": { "subtype": "article", "type": "publication", "title": "Journal article" }, "related_identifiers": [ { "scheme": "doi", "identifier": "10.5281/zenodo.4537183", "relation": "isVersionOf" } ] } }
All versions | This version | |
---|---|---|
Views | 42 | 42 |
Downloads | 25 | 25 |
Data volume | 7.1 MB | 7.1 MB |
Unique views | 29 | 29 |
Unique downloads | 23 | 23 |